RESEARCH DISCUSSION 5

 

Discussion #5

Reliability and Validity

Discuss your individual critical analysis of the posted article with in-text referencing to support your thoughts and ideas. Include an APA-formatted reference list.

Choose one of the posted articles for critique and respond to the following prompts:

1.       Identify the article chosen. Briefly describe the purpose of the study and identify the independent and dependent variables.

2.       Define the operational definition of the variables. Evaluate the operational definitions for clarity.

3.       Describe how data were collected and the instruments used to measure the data. Discuss whether these instruments were appropriate for use in this study.

4.       What steps should researchers take to ensure fidelity in data collection? Analyze what steps were taken by the researchers in this study. 

5.       Was instrument validity addressed? If it was, describe how validity was addressed. If not, discuss how the researchers might have been confident that the instrument was valid.

6.       How does the strength and quality of evidence related to reliability and validity influence the applicability of findings to clinical practice?

7.       What is your cosmic question?

ResearchEBP

Compassion Fatigue RCT

3584  |  wileyonlinelibrary.com/journal/jan J Adv Nurs. 2020;76:3584–3596.© 2020 John Wiley & Sons Ltd

Received: 2 May 2020  |  Revised: 22 July 2020  |  Accepted: 24 July 2020
DOI: 10.1111/jan.14568

O R I G I N A L R E S E A R C H : C L I N I C A L T R I A L

Effect of a compassion fatigue resiliency program on nurses’
professional quality of life, perceived stress, resilience:
A randomized controlled trial

Tuğba Pehlivan1  | Perihan Güner2

1Koç University Hospital, Istanbul, Turkey
2Istanbul Bilgi University Faculty of Health
Sciences, Istanbul, Turkey

Correspondence
Tuğba Pehlivan, Koç University Hospital,
Istanbul, Turkey.
Email: tpehlivan14@ku.edu.tr

Funding information
This research received no specific grant
from any funding agency in the public,
commercial, or not-for-profit sectors.

Abstract
Aims: This study aimed to conduct a short- and long-term Compassion Fatigue
Resiliency Program and compare its impact on nurses’ professional quality of life,
perceived stress, and resilience.
Design: The research was conducted between January 2017 January 2019 as a ran-
domized controlled trial.
Methods: The sample comprised 125 oncology–haematology nurses randomly as-
signed to a Experimental I, Experimental II, or control group. The Compassion Fatigue
Resiliency Program was conducted with the nurses as an intervention. Experimental
I received a short-term program (5 hr per day for 2 days, 10 hr in total) while
Experimental II received a long-term one (5 weeks, 2 hr per week, 10 hr in total).
Measurements were obtained during pre- and post-test and at 3-, 6-, and 12-month
follow-ups. Research hypotheses were analysed using multilevel models.
Results: The results of multilevel model analyses showed there was no statistically
significant difference between mean scores for compassion fatigue, burnout, per-
ceived stress, and resilience of nurses in the short- or long-term groups or of those
in the control group. Mean compassion satisfaction scores of nurses in the short- or
long-term groups were significantly higher than those in the control group. Mean
compassion satisfaction scores of nurses in the short- or long-term groups were sig-
nificantly higher than the control group’s pre-test mean after 6 and 12 months.
Conclusion: This study concluded that short- or long-term programs had no influ-
ence on compassion fatigue, burnout, perceived stress, and resilience; however, both
programs positively affected compassion satisfaction. We recommend that further
studies be conducted, which will help determine the effectiveness of new programs.
Impact: Caring for patients with cancer can generate work-related stress that can
negatively affect oncology nurses’ physical and emotional health, which could lead
to compassion fatigue. The program provided nurses with an opportunity to improve
their compassion satisfaction. Because there was no difference between both pro-
grams regarding mean compassion satisfaction scores, a short-term program may
be preferred to encourage more participation among nurses. We also recommend

mailto:

https://orcid.org/0000-0003-1406-5123

https://orcid.org/0000-0002-3512-296X

mailto:tpehlivan14@ku.edu.tr

http://crossmark.crossref.org/dialog/?doi=10.1111%2Fjan.14568&domain=pdf&date_stamp=2020-10-03

     |  3585PEHLIVAN ANd GÜNER

1   |   I N T R O D U C T I O N

Cancer is a chronic disease that symbolizes the limitation of con-
trol over life and death and creates many physical and psychoso-
cial problems in patients (Wang et al., 2018). The disease and its
accompanying problems challenge not only patients and families
but also the healthcare team that provides treatment and care
(Onan & Işıl, 2010). Frequent exposure to specific conditions,
such as complex cancer treatments and patient deaths, is a sig-
nificant source of stress for oncology nurses (Wilczek-Rużyczka
et al., 2019). Every day, nurses witness patients’ pain and suffering,
engage in intensive treatment, are frequently exposed to death
and dying patients, face ethical dilemmas associated with cancer
treatment, and develop close relationships with those fighting the
disease (Jones, 2017). Exposing themselves to the pain and suf-
fering, developing close relationships, and demonstrating well-de-
veloped empathy and compassion may cause nurses to experience
compassion fatigue (CF), along with decreased compassion satis-
faction (CS) (Jang et al., 2016). In addition, the increasing num-
ber of patients, excessive workload, and shortage of free time are
cited as reasons for increased job stress and burnout among on-
cology nurses according to the literature (Tuna & Baykal, 2014).
High levels of job stress negatively affect nurses’ continuance
commitment, turnover rates, patient satisfaction, and patient
safety, as well as increase burnout (Grunfeld et al., 2005). Duarte
and Pinto-Gouveia (2017) showed that approximately 25% of on-
cology nurses experienced high levels of burnout, CF, and lower
levels of CS, while Ja and Hyunjoo (2017) determined high levels
of secondary traumatic stress (STS) in 27.9%, burnout in 35% and
CS in 25.7% of nurses.

1.1 | Background

Known as a negative consequence of helping individuals experienc-
ing traumatic events or suffering, CF has an adverse effect on physi-
cal and mental health as well as on job performance and satisfaction
(Arimon-Pagès et al., 2019). Additionally, CF was found to increase
nurse turnover rates in institutions, decrease patient satisfaction,

reduce the quality of patient care, and pose a risk to patient safety
(Mooney et al., 2017). Consequently, it can be concluded that at-
tempts to prevent or reduce CF are inevitable.

According to literature examining the methods of interven-
tional studies conducted to prevent CF in oncology nurses, two
of the most effective initiatives are institutional arrangements and
training programs (Fetter, 2012; Yu et al., 2016). Yu et al. (2016)
recommended conducting specific training programs for oncology
nurses to raise awareness of possible negative effects when caring
for cancer patients. It is also necessary to increase nurse resilience
by helping them recognize and cope with CF (Rishel, 2015) and
nurses should be directed to education and training programs that
will improve their resilience (Gillman et al., 2015). The Compassion
Fatigue Resiliency Program (CFRP) is a structured, comprehensive
training program developed for nurses to prevent and minimize CF
and improve resilience. Studies examining the implementation of
CFRPs for oncology nurses (Back et al., 2014; Potter, et al., 2013;
Potter et al., 2013) showed that it was effective in preventing CF.
When interventional studies on CFRPs were examined, samples
were found to be small (Flarity et al., 2016; Pfaff et al., 2017;
Potter, et al., 2013), almost all studies were quasi-experimental,
comprising subjects assigned to the same group for pre- and post-
test (Flarity et al., 2016; Pfaff et al., 2017; Potter, et al., 2013),
and all programs were implemented for a term of 5 weeks. These
studies recommended examining the program’s effectiveness
with larger samples, particularly conducting randomized con-
trolled trials using quantitative methods and conducting a short-
term program and following-up its effectiveness for up to 1 year
(Flarity et al., 2013; Kim & Park, 2016; Pfaff et al., 2017; Potter,
et al., 2013).

2   |   T H E S T U DY

2.1 | Aims

Considering the gaps in the literature, this study aimed to implement
a CFRP for oncology nurses as a short-term and long-term program
and to compare its impact on the nurses’ professional quality of life
(CF, burnout, CS), PS, and resilience.

further studies should be conducted that include environmental improvements along
with the training programs.
Trial registered at ClinicalTrials.gov (The name of the trial register: Effect of a
Compassion Fatigue Resiliency Program; the clinical trial registration number:
NCT04372303).

K E Y W O R D S

burnout, compassion fatigue, compassion satisfaction, oncology–haematology nurses,
perceived stress, resilience

3586  |     PEHLIVAN ANd GÜNER

2.2 | Design

The research was conducted as a randomized controlled trial.

2.3 | Setting and participants

The research was conducted with nurses from the oncology–hae-
matology inpatient services, outpatient chemotherapy units, and
bone marrow transplant (BMT) units of three private hospitals in
Istanbul between January 2017 and January 2019. No sample se-
lection models were applied and of 153 nurses working in these
services, 125 nurses meeting the inclusion criteria were included
in the study. Nurses were randomly assigned to the Experimental I,
Experimental II, or control group to prevent interaction between the
subjects working in the same hospital. Of the nurses, 34 completed
the short-term program (Experimental I), 49 completed the long-
term (Experimental II), and 42 were assigned to the control group
(Figure 1).

2.3.1 | Inclusion criteria

• Nurses working in inpatient oncology–haematology, outpatient
chemotherapy, or BMT unit

2.3.2 | Exclusion criteria

• Providing care for paediatric oncology patients
• Being a nurse manager
• Not providing direct patient care

2.4 | Data collection

2.4.1 | Personal information form

A personal information form included 14 questions about socio-de-
mographics, occupational information, and working conditions, and
the level of received social support.

2.4.2 | Professional quality of life scale-IV (ProQOL-
IV)

The ProQOL-IV, developed by Stamm (2005) and adapted to Turkish
by Yeşil et al. (2010), is a self-reporting instrument consisting of 30
items and 3 subscales: CF, CS, and burnout. The scale has no total
score. Each subscale is evaluated separately. Items 1, 4, 15, 17, and
29 should be reversed scored during the evaluation. The evaluation
process of items was based on a six-digit chart ranging from “never”
(0) to “very often” (5). In the Turkish adaptation, the Cronbach’s alpha

reliability coefficients were found to be 0.83, 0.62, and 0.81 for com-
passion fatigue, burnout, and compassion satisfaction, respectively
(Yeşil et al., 2010). There was no cut-off point in the validity and reli-
ability study of the Turkish version; however, Stamm (2005) stated in
their evaluation performed with high- and low-range groups of 25%
that scores above 17 indicate high CF levels while scores below 8 in-
dicate low, scores below 18 indicate low burnout levels while scores
above 27 indicate high, and scores above 42 indicate high CS levels
while those below 33 indicate low.

2.4.3 | Perceived stress scale

The Perceived Stress Scale, developed by Cohen et al. (1983) and
adapted to Turkish by Eskin et al. (2013), uses a five-point Likert
scale with four reverse worded items (4, 5, 7, & 8) and six positively
worded items (1, 2, 3, 6, 9, & 10). There are three versions of the
scale consisting of 14, 10, and 4 items. The Cronbach’s alpha reliabil-
ity coefficient was found to be 0.82 in the Turkish adaptation of the
10-point version. It was developed to measure the degree to which
individuals perceived their life as unpredictable, uncontrollable, and
overloaded during the previous month. The lowest possible score is
0 and the highest is 40. A high total score is considered a high level
of PS.

2.4.4 | Resilience scale for adults

The Resilience Scale for Adults, developed by Friborg et al. (2005),
and adapted to Turkish by Basım and Çetin (2011) has a lowest possi-
ble score of 33 and a highest of 165. The Cronbach’s alpha reliability
coefficient of the sub-dimensions vary between 0.6-6-0.81. In addi-
tion, the value was calculated as 0.86 for the reliability of the whole
scale. Positive and negative factors are distributed on two sides of
the scale to prevent participants from prejudiced attitudes. In this
study, resilience increased in direct proportion to the scores.

2.5 | Ethical considerations

Research committee ethical approval was received from Koç
University Ethical Board of Biomedical Research on 22 December
2016 (No:2016.263.IRB2.124). Institutional permissions and verbal
and written informed consent of the nurses were obtained before
the study.

2.6 | Study procedure

This study’s principal investigator had participated online in a CFRP,
developed by Eric Gentry (Licensed Mental Health Counselor)
(2002), and received the certificate, and then conducted the pro-
gram with the nurses. Meetings were held with each institution’s

     |  3587PEHLIVAN ANd GÜNER

F I G U R E 1   CONSORT Flow Diagram showing the number of nurses in the study and those who dropped out at the different stages of the
study [Colour figure can be viewed at wileyonlinelibrary.com]

Assessed for eligibility (N = 153)

Excluded (N = 28)
Other reasons (Institution
did not plan program) (N =
28)

Lost to post-test (N = 0)
Lost to three-month follow-
up (N: 2)

Left the institution
Lost to six-month follow-up
(N: 7)

Left the institution
Lost to one-year follow-up
(N: 12)

Left the institution

Allocated to intervention
group I (N = 34)

Received the short-term
program (N = 34)

Allocated to control group
(N = 42)

Received no intervention
(N = 42)

Randomized (N = 125)

Enrollment

Allocated to intervention
group II (N = 49)

Received the long-term
program (N = 49)

Analysed at post-test (N =
34)

Allocation Allocation

Analysed at three-month
follow-up (N = 32)

Analysed at six-month
follow-up (N = 25)

Analysed at one-year
follow-up (N = 13)

Lost to post-test (N = 6)
Discontinued
intervention

Lost to three-month follow-
up (N: 0)
Lost to six-month follow-up
(N: 4)

Left the institution
Lost to one-year follow-up
(N: 8)

Left the institution

Follow-Up

Analysed at post-test (N =
43)

Lost to post-test (N = 2)
Left the institution

Lost to three-month follow-
up (N: 4)

Left the institution
Lost to six-month follow-up
(N: 7)

Left the institution
Lost to one-year follow-up
(N: 5)

Left the institution

Follow-Up

Analysed at post-test (N =
40)

Analysed at three-month
follow-up (N = 43)

Analysed at three-month
follow-up (N = 36)

Analysed at six-month
follow-up (N = 39)

Analysis Analysis

Analysed at one-year
follow-up (N = 31)

Analysed at six-month
follow-up (N = 29)

Analysed at one-year
follow-up (N = 24)

www.wileyonlinelibrary.com

3588  |     PEHLIVAN ANd GÜNER

directorate of nursing services to determine the training schedules
and content. The schedules were planned in accordance with the
hospital administration’s preferences and by taking nurses’ busy
schedules into consideration. Preliminary tests were applied to
Experimental I, Experiment II, and the control group before the
training. Experimental I received a short-term program (5 hr per
day for 2 days, 10 hr in total) while Experimental II received a long-
term one (5 weeks, 2 hr per week, 10 hr in total). No intervention
was applied to the control group. After the training, a post-test and
3-, 6-, and 12-month follow-up assessments were conducted for
all groups.

2.7 | Program

2.7.1 | Compassion fatigue resiliency program

The purpose of the program: The purpose of the program is to provide
oncology–haematology nurses with knowledge and skills that will in-
crease their level of resilience by helping them recognize compassion
fatigue, cope with its consequences, and work effectively.

2.7.2 | The objectives of the training program

Nurses, who successfully complete the program, will be able to:

• Explain the historical development of compassion fatigue among
caregivers,

• Define the developmental process of compassion fatigue,
• Specify the risk factors for compassion fatigue,
• Explain the symptoms of compassion fatigue,
• Raise awareness about their personal history,
• Explain the concept of stress and its impact on the body,
• Apply compassion fatigue resilience skills acquired in the program,
• Professionally create a self-directed resilience plan.

2.8 | Validity and reliability

The included scales have been tested for validity and reliability for
various settings and countries (Çolak Sarı, 2018; Deible et al., 2015;
Rushton et al., 2015; Tarantino et al., 2013; Tekin, 2011; Traeger
et al., 2013). We chose three private hospitals in Istanbul, which are
considered to be close to each other, such as institution (etc. work-
ing conditions) and the socio-demographic and professional char-
acteristics of nurses (age, educational status, clinical experience,
willingness to work in oncology, voluntarily career choice, and etc.),
that are known to affect outcome variables. Then, we randomly as-
signed each hospital to the Experimental I, Experimental II, or con-
trol group to prevent interaction between the nurses working in the
same hospital. Finally, the program was conducted by the principal
investigator who had participated online in a CFRP and received the
certificate.

2.9 | Data analysis

SPSS 25.0 software package was used to analyse data. Prior to the
employment of parametric statistical techniques, the Shapiro–Wilk
test was used to ensure the data distributed normally. A chi-square
test and one-way ANOVA determined whether the scales and in-
formation on demographic characteristics and work environment
differed among pre-initiative groups; in other words, to measure ho-
mogeneity. There was a statistically significant difference between
the groups regarding age, total work hours per week, total years of
professional experience, total years of oncology experience, level of
received social support, and marital status (p < .05). The comparison of the mean scores for nurses’ compassion fatigue, burnout, compas- sion satisfaction, perceived stress, and resilience according to pre- initiative groups is given in Table 1. Variables found to be statistically significant were included in the model as a correction factor for primary hypotheses. Primary research hypotheses were analysed using multilevel models (MLM). Variables Experimental I (N= 34) Mean (SD) Experimental II (N= 49) Mean (SD) Control (N= 42) Mean (SD) p-Value Compassion fatigue 15.2 (7.7) 12.4 (6.0) 13.5 (8.1) .198 Burnout 21.3 (6.7) 19.2 (6.8) 20.3 (6.8) .382 Compassion satisfaction 38.5 (7.4) 39.6 (6.7) 35.5 (7.8) .036* Perceived stress 32.7 (3.1) 31.8 (3.6) 31.7 (3.5) .308 Resilience 139.4 (15.6) 133.3 (19.3) 134.7 (16) .261 Abbreviations: Experimental II, Long-term training group. †Standard deviations are given in parentheses. ‡Experimental I, Short-term training group. §p-Values indicate the statistically significant differences between groups. ¶Bold values indicate statistically significant values; SD: standard deviation. *Post hoc analysis indicated a statistically significant difference between Experimental II and control group. T A B L E 1   Comparison of the baseline mean scores of compassion fatigue, burnout, compassion satisfaction, perceived stress, and resilience among groups (N: 125)      |  3589PEHLIVAN ANd GÜNER This analysis method has some advantages over others, such as repeated measures ANOVA. These advantages can be defined as the ability to deal with missing data in MLM and be included in the analysis process of investigating individual differences (ran- dom effect) (Tab achnick & Fidell, 2015). Consequently, MLM was used to analyse the data. In the analysis process, distribution of the random intercepts and slopes was determined, and models were compared using log-likelihood function based on different correlation assumptions for repeated measures and random ef- fects. Then, the effect of group, time, and group X time interac- tions and other explanatory variables were included in the model and the final model-based results were reported. In the process of T A B L E 2   Baseline characteristics of nurses (N: 125) Characteristics Experimental I (N= 34) N (%) Experimental II (N= 49) N (%) Control (N= 42) N (%) Educational status Vocational high school 7 (14.3) 5 (14.7) 5 (11.9) Associate's degree 2 (4.1) 4 (11.8) 3 (7.1) Bachelor's degree 35 (71.4) 21 (61.8) 32 (76.2) Postgraduate 5 (10.2) 4 (11.8) 2 (4.8) Level of received social support Enough 10 (20.4) 13 (38.2) 20 (47.6) Not enough 39 (79.6) 21(61.8) 22 (52.4) Willing to work in oncology Yes 47 (95.9) 34 (100) 40 (95.2) No 2 (4.1) 0 (0) 2 (4.8) Voluntary career choice Yes 38 (77.6) 27 (79.4) 28 (66.7) No 11 (22.4) 7 (20.6) 14 (33.3) Type of shift Permanent day shift 8 (16.3) 11 (32.4) 12 (28.6) Day-night shift 41 (83.7) 23 (67.6) 30 (71.4) Gender Female 44 (89.8) 32 (94.1) 36 (85.7) Male 5 (10.2) 2 (5.9) 6 (14.3) Marital status Married 14 (28.6) 2 (5.9) 8 (19) Single 35 (71.4) 32 (94.1) 34 (81) Department Oncology inpatient service 7 (14.3) 8 (23.5) 11 (26.2) Haematology inpatient service 11 (22.4) 3 (8.8) 0 (0) Oncology–haematology inpatient service 0 (0) 4 (11.8) 21 (50.0) Bone marrow transplantation unit 24 (49.0) 12 (35.3) 0 (0) Outpatient chemotherapy unit 7 (14.3) 7 (20.6) 10 (23.8) Experimental I Mean (SD) Experimental II Mean (SD) Control Mean (SD) Age 25.0 (5.1) 27.8 (5.3) 27.2 (5.3) Work hours per week 53.8 (4.7) 48.1 (3.8) 48.4 (1.6) Work hours per day 11.1 (1.0) 11.5 (1.0) 11.5 (1.0) Total years of nursing experience 3.0 (3.7) 5.6 (5.7) 4.9 (4.7) Total years in the current department 1.9 (1.8) 2.9 (2.2) 3.3 (2.8) Abbreviation: Experimental II, Long-term training groupSD, standard deviation. †Percentage (%) values and standard deviations are given in parentheses. ‡Experimental I, Short-term training group. 3590  |     PEHLIVAN ANd GÜNER determining the most significant model, variables with a p-value greater than 0.10 were removed, respectively, and variables with a p-value of approximately 0.10 or less were kept in the model, while p < .05 was considered significant. 3   |   R E S U LT S 3.1 | Demographics Baseline characteristics are given in Table 2. 3.2 | Multilevel model results of CF subscale There was no statistically significant difference between the main effect of the initiative (group) variable and the outcome variable of CF (p = .541), indicating that there was no statistically significant difference between groups regarding mean CF scores of nurses receiving short-term training (mean = 16.43, SD 8.46), long-term training (mean = 14.20, SD 8.17), or those in the control group (mean = 14.93, SD 8.04). The effect of the time variable related to the CF outcome variable was statistically significant (p < .001). Further analysis of the time variable indicated that the 3-month (mean = 15.45, SD 9.21), 6-month (mean = 16.49, SD 8.97), and 1-year (mean = 14.90, SD 8.05) follow-up scores of nurses were greater than their pre-test mean scores (mean = 13.54, SD 7.25) and the difference was statistically significant (Table 3). The group X time interaction for the CF outcome variable was statistically significant (p = .047). There was a statistically significant differ- ence between groups over the time period regarding mean CF scores. The difference between pre- and post-test measurements in the group that received only long-term training was statistically significant regarding the difference between pre- and post-test measurements for the control group (Table 3). 3.3 | Multilevel model results of burnout subscale There was no statistically significant difference between the main effect of the initiative (group) variable and the burnout variable (p = .227), indicating that there was no statistically significant dif- ference between the groups regarding mean burnout scores of nurses receiving short-term training (mean = 22.30, SD 7.22), long-term training (mean = 20.05, SD 6.90), or those in the control group (mean = 21.01, SD 6.89). The effect of the time variable re- lated to the burnout outcome variable was statistically significant (p = .009). Further analysis of the time variable indicated that 3-month (mean = 21.26, SD 7.45), 6-month (mean = 21.85, SD 6.97), and 1-year (mean = 21.51, SD 7.19) follow-up scores of nurses were greater than their pre-test mean scores (mean = 20.14, SD 6.77) and the difference was statistically significant (Table 4). The group X time interaction for the burnout outcome variable was not statistically significant (p = .710). There was no statistically significant differ- ence between groups over the time period regarding mean burnout scores. 3.4 | Multilevel model results of CS subscale A statistically significant difference was found between the main effect of the initiative (group) variable and the CS outcome variable (p < .001), indicating that there was a statistically significant differ- ence between all groups regarding mean CS scores. Further analysis of T A B L E 3   Multilevel analysis with compassion fatigue as outcome variable β SE [95% CI] p Intercept 1.89 3.68 [−5.40, 9.17] .609 Experimental I versus. Control 0.86 1.85 [−2.78, 4.50] .641 Experimental II versus. Control −1.66 1.71 [−5.04, 1.72] .335 Posttest 0.40 0.97 [−1.51, 2.32] .678 Three-month follow-up 2.48 1.14 [0.22, 4.73] .031 Six-month follow-up 3.16 1.27 [0.67, 5.66] .013 One-year follow-up 3.83 1.37 [1.12, 6.54] .006 Marital status 3.83 1.59 [0.68, 6.98] .017 Level of received social support 3.10 1.32 [0.49, 5.71] .021 Post-test Experimental I versus. Control 0.42 1.44 [−2.41, 3.25] .771 Experimental II versus. Control 3.16 1.35 [0.51, 5.81] .020 Three-month follow-up Experimental I versus. Control −0.97 1.68 [−4.28, 2.34] .564 Experimental II versus. Control −0.71 1.56 [−3.78, 2.37] .650 Six-month follow-up Experimental I versus. Control −1.05 1.87 [−4.73, 2.64] .577 Experimental II versus. Control 0.61 1.70 [−2.73, 3.96] .718 One-year follow-up Experimental I versus. Control −3.31 2.22 [−7.68, 1.07] .138 Experimental II versus. Control −2.37 1.84 [−6.01, 1.26] .199 Note: Experimental II, Long-term training group. ‡Bold values indicate statistically significant values. §Control was the reference group; SE, standard error. †Experimental I, Short-term training group.      |  3591PEHLIVAN ANd GÜNER the group variable indicated that mean scores of nurses who received short-term (mean = 37.84, SD 8.05) and long-term (mean = 39.49, SD 7.14) training were greater than the control group (mean = 33.84, SD 8.85) and the difference was statistically significant (Table 5). The effect of the time variable related to CS outcome variable was statistically significant (p = .013). Further analysis of the time vari- able indicated that the 1-year (mean = 36.06, SD 9.73) and 6-month (mean = 36.32, SD 9.76) follow-up scores of nurses were less than their pre-test mean scores (mean = 37.92, SD 7.40) and the difference was statistically significant (Table 5). The group X time interaction for the CS outcome variable was statistically significant (p = .23). There was a statistically significant difference between groups over the pe- riod regarding mean CS scores. The difference between the pre-test and 6-month and the pre-test and 1-year measurements of nurses re- ceiving the short- or long-term training was statistically significantly higher than the pre-test and 6-month measurements and the pre-test and 1-year measurements of nurses in the control group (Table 5). 3.5 | Multilevel model results of PS scale There was no statistically significant difference between the main effect of the initiative (group) variable and PS (p = .849). This results indicated no statistically significant difference between the groups regarding mean PS scores of nurses receiving short-term training (mean = 32.29, SD 3.27), long-term training (mean = 31.55, SD 3.28), and those in the control group (mean = 31.66, SD 3.96). The effect of the time variable related to PS outcome variable was not statistically significant (p = .742). The group X time interaction for the PS outcome variable was not statis- tically significant (p = .510). There was no statistically significant differ- ence in mean PS scores between groups over the period. 3.6 | Multilevel model results of the resilience scale There was no statistically significant difference between the main effect of the initiative (group) variable and resilience (p = .625). T A B L E 4   Multilevel analysis with burnout as outcome variable β SE [95% CI] p Intercept −15.40 8.47 [−32.18, 1.38] .072 Experimental I versus. Control −2.16 1.55 [−5.22, 0.90] .165 Experimental II versus. Control −1.78 1.23 [−4.22, 0.66] .151 Post-test 0.49 0.50 [−0.48, 1.47] .321 Three-month follow-up 1.11 0.59 [−0.04, 2.27] .050 Six-month follow-up 2.15 0.65 [0.88, 3.42] <.001 One-year follow-up 2.32 0.73 [0.88, 3.76] .002 Level of received social support 2.07 1.08 [−0.08, 4.22] .058 Work hours per week 0.45 0.16 [0.13, 0.77] .006 Type of shift 1.32 0.72 [−0.10, 2.74] .069 Marital status 2.75 1.32 [0.13, 5.36] .040 Department 0.88 0.45 [−0.01, 1.77] .053 Abbreviation: Experimental II, Long-term training group. ‡Bold values indicate statistically significant values. §Control was the reference group; SE, standard error. †Experimental I, Short-term training group. T A B L E 5   Multilevel analysis with compassion satisfaction as outcome variable β SE [95% CI] p Experimental I versus. Control 5.60 1.82 [2.01, 9.19] .002 Experimental II versus. Control 4.59 1.51 [1.62, 7.56] .003 Post-test 0.00 0.91 [−1.78, 1.78] .999 Three-month follow-up −0.96 1.07 [−3.06, 1.13] .366 Six-month follow-up −4.43 1.21 [−6.81, −2.06] <.001 One-year follow-up −5.89 1.34 [−8.52, −3.26] <.001 Level of received social support −2.63 1.22 [−5.03, −0.23] .032 Work hours per week −0.46 0.16 [−0.78, −0.13] .006 Posttest Experimental I versus. Control −0.18 1.34 [−2.81, 2.46] .895 Experimental II versus. Control 0.45 1.26 [−2.02, 2.92] .720 Three-month follow-up Experimental I versus. Control −0.55 1.57 [−3.63, 2.53] .724 Experimental II versus. Control 0.14 1.46 [−2.71, 3.00] .921 Six-month follow-up Experimental I versus. Control 4.20 1.79 [0.69, 7.71] .019 Experimental II versus. Control 3.75 1.62 [0.56, 6.93] .021 One-year follow-up Experimental I versus. Control 5.56 2.16 [1.31, 9.80] .010 Experimental II versus. Control 5.12 1.80 [1.59, 8.65] .005 Abbreviation: Experimental II, Long-term training group. ‡Bold values indicate statistically significant values. §Control was the reference group; SE, standard error. †Experimental I, Short-term training group. 3592  |     PEHLIVAN ANd GÜNER This result indicated that there was no statistically significant dif- ference between the groups regarding mean resilience scores of nurses in Experimental I (mean = 135.18, SD 17.45), Experimental II (mean = 131.62, SD 19.08), and control group (mean = 133.36, SD 16.59). The effect of the time variable related to the resilience out- come variable was statistically significant (p = .005). Further analy- sis indicated that the 3-month (mean = 131.15, SD 17.23), 6-month (mean = 131.17, SD 19.55), and 1-year (mean = 130.82, SD 20.02) follow-up scores of nurses were less than the pre-test mean scores (mean = 135.43, SD 17.29) and the difference was statistically sig- nificant (Table 6). The group X time interaction for the resilience outcome variable was not statistically significant (p = .510). There was no statistically significant difference in mean resilience scores between groups over the period. 4   |   D I S C U S S I O N This study was the first to be conducted as a randomized controlled trial of the effects of short- and long-term CFRP on CF, burnout, CS, PS, and resilience. 4.1 | Results Obtained from the ProQOL Scale In our study, both the short- and long-term CFRPs had no effect on the nurses’ mean CF scores. Contrary to our study, some con- ducted on a long-term training program (Back et al., 2014; Potter, et al., 2013) showed that mean CF scores during the pre-training period were statistically greater than those measured immediately afterward and at the 6-month follow-up, while another study (Pfaff et al., 2017) suggested no difference in mean CF scores. In addition, other studies (Flarity et al., 2013; Flarity, et al., 2016) showed, con- trary to our study, that there was a statistically significant decrease in nurses’ mean CF scores measured immediately after the CFRP. In our study, the mean CF scores of nurses receiving long-term training were found to have increased immediately afterward. In our study, contrary to other studies (Flarity, et al., 2016; Potter, et al., 2013), the reason that there was no significant change in nurses’ mean CF scores may be associated with low mean scores obtained at the begin- ning of the study. Additionally, total years of nursing experience was considerably low (nurses receiving short-term training (mean = 1.9, SD 1.8); long-term training (mean = 2.9, SD 2.2); and control group (mean = 3.3, SD 2.8)). Consequently, another reason that nurses’ CF scores increased immediately afterward may be related to their en- hanced CF awareness following the training. Our study suggests that the gradual increase in mean CF scores may also be related to nurses’ heavy workloads, including time-consuming orientation and training to newly hired nurses and other negative factors affecting turnover rates in institutions, which was observed to be at high levels during the research period. In fact, 45.6% of nurses left their institutions in the year following this study. In our study, both the short- and long-term CFRPs had no influence on mean burnout scores. While this result is consistent with a study (Pfaff et al., 2017), others (Flarity et al., 2013; Potter et al., 2013) sug- gest the opposite. Two studies (Flarity et al., 2013; Potter et al., 2013) showed that nurses’ mean burnout scores indicated a statistically sig- nificant decrease after a 5-week CFRP; however, Pfaff et al. (2017) showed no significant difference after a 6-week CFRP. In our study, the reason that no significant change was found in mean burnout scores, as in CF, may be related to low mean scores obtained at the beginning of the study. In addition, limited work experience (nurses receiving short- term training (mean = 1.9, SD 1.8); long-term training (mean = 2.9, SD 2.2); control group (mean = 3.3, SD 28)) may explain the low mean scores for burnout. Another reason that the mean burnout scores in- creased gradually may be related to organizational changes in hospi- tals. Nurses assigned to public institutions during the study period may have experienced an increase in work environment stressors such as a heavy workload or reduced staffing. Recent studies (Pfaff et al., 2017; Potter et al., 2013) performed evaluations immediately after the pro- grams and thus long-term follow-up results could not be compared. The results are not consistent with each other, so there is no certainty of the long-term impact of CFRP on burnout mean scores. Our study showed that the mean CS scores of nurses receiv- ing either the short- or long-term training were greater than those in the control group. In a study by Potter et al. (2013) evaluating a 5-week CFRP, the nurses experienced high levels of CS afterward. Interestingly, our study indicated that mean scores for CS did not change in either group immediately afterward, while they increased at the 6-month and 1-year follow-ups. It was also observed that the number of nurses leaving the institutions increased over the period. T A B L E 6   Multilevel analysis with resilience as outcome variable β SE [95% CI] p Intercept 181.27 10.18 [161.11, 201.43] <.001 Experimental I versus. Control 2.24 3.31 [−4.32, 8.81] .500 Experimental II versus. Control −0.92 3.08 [−7.02, 5.18] .765 Posttest 0.18 1.33 [−2.44, 2.79] .894 Three-month follow-up −3.91 1.51 [−6.88, −0.94] .010 Six-month follow-up −4.43 1.63 [−7.64, −1.23] .007 One-year follow-up −4.59 1.83 [−8.19, −1.00] .013 Gender −9.36 4.09 [−17.47, −1.25] .024 Marital status −10.41 3.76 [−17.86, −2.96] .007 Type of shift −0.52 0.29 [−1.11, 0.06] .077 Level of received social support −8.78 2.73 [−14.18, −3.38] .002 Abbreviation: Experimental II, Long-term training group. ‡Bold values indicate statistically significant values. §Control was the reference group; SE, standard error. †Experimental I, Short-term training group.      |  3593PEHLIVAN ANd GÜNER This result might be related to the fact that nurses with higher mean scores for CS preferred to stay in their institution. In a study conducted by Güner et al. (2018), a low level of CS was one of the main reasons that oncology nurses would leave their institution, an- swering “yes” to the question: “Would you leave oncology unit now if you had a chance?” In addition, because there was no long-term follow-up in the study conducted by Potter et al. (2013), only the measurements obtained immediately after the program were com- pared. Consequently, there is no sufficient evidence to prove long- term impact of CFRPs and their effect on mean CS scores is unclear according to the results of our study and literature review. 4.2 | Results obtained from the PS Scale Our study showed that the short- and long-term CFRPs had no influ- ence on the mean PS scores. However, Pfaff et al. (2017) showed that mean clinical stress scores of oncology nurses participating in a 6-week CFRP showed a statistically significant decrease afterward. The mean PS scores of nurses in both groups were also found to fluctuate. This result can be associated with the definition of the PS variable. It is known that the effect of events considered to cause stress depends on the individual assessment of the particular event as a stress factor and perception may vary depending on the situation (Yerlikaya, 2009). In this context, the concept of “PS” stands out, referring to the level of stress individuals experience and its assessment as overloaded and uncontrollable (Cohen et al., 1983). Consequently, PS levels reflect the perception of subjective stress. At this point, subjective stress perception and its influencing factors, depending on how individuals perceived the determinants, may have led to fluctuations in the mean PS scores of nurses receiving the short- or long-term training. A nurse in our study who participated in the long-term training, for instance, noted on the PS scale at the 3-month follow-up that: “I filled out the questionnaire under the influence of premenstrual syndrome and needed to indicate that it may affect the results of your study.” 4.3 | Results obtained from the resilience scale for adults Our study showed that the short- and long-term CFRPs had no influ- ence on mean scores of resilience and, in fact, the scores decreased in the two groups. Lowe (2013) stated that resilience consists of two components – individual and environmental factors – and that inter- action between the two improves the resilience level. The effect of individual characteristics and environmental factors on the devel- opment of resilience is undeniable. Consequently, in our study both short- and long-term CFRPs may have had no influence on nurses’ mean resilience scores because no environmental improvements were performed in the current program. In the literature, interven- tional studies aimed to improve the resilience of oncology nurses, indicated that institutional support plays an important role in improv- ing the resilience level (Fitch et al., 2006; Sandgren et al., 2006). The concept of resilience is also defined as an innate energy or motivating life force (Waite & Richardson, 2004). Based on these definitions, de- veloping resilience is not easy to achieve in a short time and this can be associated with the insufficient outcomes of our study and the cur- rent training program. In our study, the mean CF and burnout scores increased unexpectedly. This may have caused nurses to experience weakness that prevented them from developing resilience. 4.4 | Study limitations There are some limitations to this research. First, the fact that nurses left their institutions contributed to the low number of participants at the one-year follow-up. Another limitation that may have affected the outcome of the study was that some nurses working in private hospitals made a transfer to public hospitals, which coincided with the period when the research was conducted. The Turkish Ministry of Health periodically opens employment opportunities for staff nurses in state hospitals and nurses often prefer to work in pub- lic hospitals. Suddenly, the increased workload in the private hos- pitals, where the number of nurses decreased significantly, may have negatively affected the outcome of the study. Therefore, the effectiveness of the CFRP program can be concluded more pre- cisely by repeating this first study. The fact that the hospitals where the research was conducted had different characteristics (working conditions, number of nurses, and facilities) may have also affected outcome variables. Also, we recommend the study be repeated in institutions with different characteristics, which may lead to differ- ent results. Finally, the collected study data are based solely on the nurses’ self-reporting. 5   |   C O N C L U S I O N This study, which attempted to determine the effects of a short- and long-term CFRP on CF, burnout, CS, PS, and resilience, showed that neither the short- nor long-term training had any influence on CF, burn- out, PS, and resilience; however, both methods were found to posi- tively affect CS levels. Because there was no difference between both programs regarding mean CS scores, a short-term program may be pre- ferred to encourage more participation among nurses. The results also showed that neither the short- nor long-term CFRP training for nurses had any influence on the mean scores for resilience and PS. Although nurses receiving the training reported that the participant-centred program had positive contributions, we recommend that further stud- ies be conducted as randomized controlled trials and include 1-year follow-ups to determine the effectiveness of a program implemented for the first time. While the results did show that CFRP had a positive effect on CS, neither the short- nor long-term programs showed ef- fect on CF, burnout, PS, and resilience and thus oncology–haematol- ogy nurses would not likely benefit from such a program, or would it be cost effective. However, data obtained after new studies are con- ducted may offer more specific findings about whether conducting 3594  |     PEHLIVAN ANd GÜNER this program with oncology–haematology nurses is beneficial or not. While the literature indicates institutions should also take actions to prevent CF and enhance resilience (Fetter, 2012; Yu et al., 2016), one nurse participating in our study said: “Changing our perspective is okay, but there is definitely a limit and it's hard to maintain. Institutional actions also need to be taken.” Therefore, the impact of organizational factors is also undeniable to prevent CF and improve resilience and we rec- ommend further studies be conducted that include organizational ar- rangements along with the training programs. Finally, conducting more qualitative studies is recommended for in-depth measurement of such social impacts as the perceptions and recommendations of the nurses, as well as the training method of the program and characteristics of the educators, which cannot be measured by quantitative studies. A C K N O W L E D G E M E N T S The authors would like to thank all nurses who participated in this study for their valuable contributions. This study was presented as a poster in the EONS 12 at the ESMO 2019 conference, Barcelona, 27 September–1 October 2019. An abstract was published in Annals of Oncology, 30 (5), v847. C O N F L I C T S O F I N T E R E S T No conflict of interest has been declared by the author(s). A U T H O R C O N T R I B U T I O N S All authors have agreed on the final version and meet at least one of the following criteria (recommended by the ICMJE): substantial contri- butions to conception and design, acquisition of data, or analysis and interpretation of data; drafting the article or revising it critically for important intellectual content. T Pehlivan, P Güner: Made substantial contributions to conception and design, or acquisition of data, or anal- ysis and interpretation of data; T Pehlivan, P Güner: Involved in draft- ing the manuscript or revising it critically for important intellectual content; T Pehlivan, P Güner: Given final approval of the version to be published. Each author should have participated sufficiently in the work to take public responsibility for appropriate portions of the con- tent; T Pehlivan, P Güner: Agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. P E E R R E V I E W The peer review history for this article is available at https://publo ns.com/publo n/10.1111/jan.14568. O R C I D Tuğba Pehlivan https://orcid.org/0000-0003-1406-5123 Perihan Güner https://orcid.org/0000-0002-3512-296X R E F E R E N C E S Arimon-Pagès, E., Torres-Puig-Gros, J., Fernández-Ortega, P., & Canela- Soler, J. (2019). Emotional impact and compassion fatigue in oncology nurses: Results of a multicentre study. European Journal of Oncology Nursing, 43, 1–6. https://doi.org/10.1016/j.ejon.2019.09.007 Back, A. L., Deignan, P. F., & Potter, P. A. (2014). Compassion, compas- sion fatigue and burnout: Key insight for oncology professionals. The ASCO Educational Book, 454–459. https://doi.org/10.14694/ EdBook_AM.2014.34.e454 Basım, H. N., & Çetin, F. (2011). Yetişkinler için psikolojik dayanıklılık ölçeğinin güvenilirlik ve geçerlilik çalışması. Türk Psikiyatri Dergisi, 22(2), 104–114. Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress. Journal of Health and Social Behaviour, 24, 385–396. https://doi.org/10.2307/2136404 Çolak Sarı, E. (2018). Riskli birimlerde çalışan hemşirelerin psikolojik sağlamlıklarının değerlendirilmesi. [Yayımlanmamış yüksek lisans tezi]. Avrasya Üniversitesi Sağlık Bilimleri Enstitüsü Hemşirelik Ababilim Dalı. Deible, S., Fioravanti, M., Tarantino, B., & Cohen, S. (2015). Implementation of an integrative coping and resiliency program for nurses. Global Advances in Health and Medicine, 4(1), 28–33. https:// doi.org/10.7453/gahmj.2014.057 Duarte, J., & Pinto-Gouveia, J. (2017). Empathy and feelings of guilt ex- perienced by nurses: A cross-sectional study of their role in burn- out and compassion fatigue symptoms. Applied Nursing Research, 35, 42–47. https://doi.org/10.1016/j.apnr.2017.02.006 Eskin, M., Harlak, H., Demirkıran, F., & Dereboy, Ç. (2013). Algılanan stres ölçeğinin Türkçe’ye uyarlanması: Güvenirlik ve geçerlik analizi. New/ Yeni Symposium Journal, 51(3), 132–140. Fetter, K. L. (2012). We grieve too: One inpatient oncology unit’s in- terventions for recognizing and combating compassion fatigue. Clinical Journal of Oncology Nursing, 16(6), 559–561. https://doi. org/10.1188/12.CJON.559-561 Fitch, M. I., Matyas, Y., & Robinette, M. (2006). Caring for the caregivers: Innovative program for oncology nurses. Canadian Oncology Nursing Journal, 16(2), 110–122. https://doi.org/10.5737/11819 12x16 2110115 Flarity, K., Gentry, E., & Mesnikoff, N. (2013). The effectivenenss of an educational program on preventing and treating compassion fatigue in emergency nurses. Advanced Emergency Nursing Journal, 35(3), 247–258. https://doi.org/10.1097/TME.0b013 e3182 9b726f Flarity, K., Nash, K., Jones, W., & Steinbruner, D. (2016). Intervening to improve compassion fatigue resiliency in forensic nurses. Advanced Emergency Nursing Journal, 38(2), 147–156. https://doi.org/10.1097/ TME.00000 00000 000101 Flarity, K., Rhodes, W. J., & Reckard, P. (2016). Intervening to improve compassion fatigue resiliency in nurse residents. Journal of Nursing Education and Practice, 6(12), 99–104. https://doi.org/10.5430/jnep. v6n12p99 Friborg, O., Barlaug, D., Martinussen, M., Rosenvinge, J. H., & Hjemdal, O. (2005). Resilience in relation to personality and intelligence. International Journal of Methods in Psychiatric Research, 14(1), 29–42. https://doi.org/10.1002/mpr.15 Gentry, E., Baranowsky, A., & Dunning, K. (2002). The Accelerated Recovery Program for compassion fatigue. In C. R. Figley (Ed.), Treating compassion fatigue (pp. 123–138). Brunner/Mazel. Gillman, L., Adams, J., Kovac, R., Kilcullen, A., House, A., & Doyle, C. (2015). Strategies to promote coping and resilience in oncology and palliative care nurses caring for adult patients with malignancy: A comprehensive systematic review. JBI Database of Systematic Reviews & Implementation Reports, 13(5), 131–204. https://doi.org/10.11124/ jbisr ir-2015-1898 Grunfeld, E., Zitzelsberger, L., Coristine, M., Whelan, T. J., Aspelund, F., & Evans, W. K. (2005). Job stress and job satisfaction of cancer care workers. Psycho-oncology, 14(1), 61–69. https://doi.org/10.1002/ pon.820 Güner, P., Kocaman Yıldırım, N., İnci, F., Hiçdurmaz, D., Santiago Fernandez, R., Özdemir, S., İnce, A., & Yıldırım, Y. (2018). Onkoloji alanında çalışan hemşirelerin psikososyal bakım verme konusundaki gereksinimlerinin belirlenmesi. Proje Çalışması. https://publons.com/publon/10.1111/jan.14568 https://publons.com/publon/10.1111/jan.14568 https://orcid.org/0000-0003-1406-5123 https://orcid.org/0000-0003-1406-5123 https://orcid.org/0000-0002-3512-296X https://orcid.org/0000-0002-3512-296X https://doi.org/10.1016/j.ejon.2019.09.007 https://doi.org/10.14694/EdBook_AM.2014.34.e454 https://doi.org/10.14694/EdBook_AM.2014.34.e454 https://doi.org/10.2307/2136404 https://doi.org/10.7453/gahmj.2014.057 https://doi.org/10.7453/gahmj.2014.057 https://doi.org/10.1016/j.apnr.2017.02.006 https://doi.org/10.1188/12.CJON.559-561 https://doi.org/10.1188/12.CJON.559-561 https://doi.org/10.5737/1181912x162110115 https://doi.org/10.1097/TME.0b013e31829b726f https://doi.org/10.1097/TME.0000000000000101 https://doi.org/10.1097/TME.0000000000000101 https://doi.org/10.5430/jnep.v6n12p99 https://doi.org/10.5430/jnep.v6n12p99 https://doi.org/10.1002/mpr.15 https://doi.org/10.11124/jbisrir-2015-1898 https://doi.org/10.11124/jbisrir-2015-1898 https://doi.org/10.1002/pon.820 https://doi.org/10.1002/pon.820      |  3595PEHLIVAN ANd GÜNER Ja, K., & Hyunjoo, N. (2017). A study of the relationships between com- passion fatigue, compassion satisfaction, depression, anxiety and sleep disorders among oncology nurses. Asian Oncology Nursing, 17(2), 116–123. https://doi.org/10.5388/aon.2017.17.2.116 Jang, I., Kim, Y., & Kim, K. (2016). Professionalism and professional qual- ity of life for oncology nurses. Journal of Clinical Nursing, 25, 2835– 2845. https://doi.org/10.1111/jocn.13330 Jones, A. K. (2017). Oncology nurse retreat: A strength-based approach to self-care and personal resilience. Clinical Journal of Oncology Nursing, 21(1), 259–262. https://doi.org/10.1188/17.CJON.259-262 Kim, Y. A., & Park, J. S. (2016). Development and application of an over- coming compassion fatigue program for emergency nurses. Journal of Korean Academy of Nursing, 46(2), 260–270. https://doi.org/10.4040/ jkan.2016.46.2.260 Lowe, L. D. (2013). Creating a caring work environment and fostering nurse resilience. International Journal of Human Caring, 17, 52–59. https://doi.org/10.20467/ 1091-5710.17.4.52 Mooney, C., Fetter, K., Gross, B. W., Rinehart, C., Lynch, C., & Rogers, F. B. (2017). A preliminary analysis of compassion satisfaction and com- passion fatigue with considerations for nursing unit specialization and demographic factors. Journal of Trauma Nursing, 24(3), 158–163. https://doi.org/10.1097/JTN.00000 00000 000284 Onan, N., & Işıl, Ö. (2010). Onkoloji birimlerinde çalışan hemşirelerde stres, tükenmişlik ve başa çıkma: Literatür gözden geçirme. Maltepe Üniversitesi Hemşirelik Bilim Ve Sanatı Dergisi Sempozyum Özel Sayısı, 264-271. Pfaff, K. A., Freeman-Gibb, L., Patrick, L. J., DiBiase, R., & Moretti, O. (2017). Reducing the “cost of caring” in cancer care: Evaluation of a pilot interprofessional compassion fatigue resiliency pro- gramme. Journal of Interprofessional Care, 31(4), 512–519. https://doi. org/10.1080/13561 820.2017.1309364 Potter, P., Deshields, T., Berger, J. A., Clarke, M., Olsen, S., & Chen, L. (2013). Evaluation of a compassion fatigue resiliency programme for oncology nurses. Oncology Nursing Forum, 40(2), 180–187. https:// doi.org/10.1188/13.ONF.180-187 Potter, P., Deshields, T., & Rodriguez, S. (2013). Developing a systemic program for compassion fatigue. Nursing Administration Quarterly, 20(1), 326–332. https://doi.org/10.1097/NAQ.0b013 e3182 a2f9dd Rishel, C. J. (2015). The role of resilience and mindful leadership in on- cology nursing. Oncology Nursing Forum, 42(2), 198–199. https://doi. org/10.1188/15.ONF.198-199 Rushton, C. H., Batcheller, J., Schroeder, K., & Donohue, P. (2015). Burnout and resilience among nurses practicing in high-intensity set- tings. American Journal of Critical Care, 24(5), 412–421. https://doi. org/10.4037/ajcc2 015291 Sandgren, A., Thulesius, H., Fridlund, B., & Petersson, K. (2006). Striving for emotional survival in palliative cancer nursing. Qualitative Health Research, 16(1), 79–96. https://doi.org/10.1177/10497 32305 283930 Stamm, B. H. (2005). The ProQOL Manual. The Professional Quality of Life Scale: Compassion satisfaction, burnout & compassion fatigue/ secondary trauma scales. Sidran Press; s:1–13. Tabachnick, B. G., & Fidell, L. S. (2015). Çok değişkenli istatistiklerin kullanımı. (Çev. Ed. Mustafa Baloğlu), (6. Basım). Ankara: Nobel Akademik Yayıncılık Eğitim Danışmanlık. Tarantino, B., Earley, M., Audia, D., D’Adamo, C., & Berman, B. (2013). Qualitative and quantitative evaluation of a pilot integrative coping and resiliency program for healthcare professionals. Explore, 9(1), 44–47. https://doi.org/10.1016/j.explo re.2012.10.002 Tekin, E. (2011). Askeri hastanelerde çalışan hemşirelerin psikolojik dayanıklılık ve tükenmişlik düzeylerinin belirlenmesi. [Yayımlanmamış yüksek lisans tezi]. Gazi Üniversitesi Sağlık Bilimleri Enstitüsü Hemşirelik Anabilim Dalı. Traeger, L., Park, E. R., Sporn, N., Repper-DeLisi, J., Convery, M. S., Jacobo, M., & Pirl, W. F. (2013). Development and evaluation of tar- geted psychological skills training for oncology nurses in managing stressful patient and family encounters. Oncology Nursing Forum, 40(4), E327–E336. https://doi.org/10.1188/13.ONF.E327-E336 Tuna, R., & Baykal, Ü. (2014). The relationship between job stress and burnout levels of oncology nurses. Asia-Pacific Journal of Oncology Nursing, 1(1), 33–39. https://doi.org/10.4103/2347-5625.135818 Waite, P. J., & Richardson, G. E. (2004). Determining the effiacy of resil- iency training. Journal of Allied Health, 33, 178–183. Wang, T., Molassiotis, A., Chung, B. P. M., & Tan, J. (2018). Unmet care needs of advanced cancer patients and their informal caregivers: A systematic review. BMC Palliative Care, 17(96), 2–29. https://doi. org/10.1186/s1290 4-018-0346-9 Wilczek-Rużyczka, E., Dębska, G., Pasek, M., & Zwierzchowska, M. (2019). The mediational effect of coherence on the relationship between mental load and job burnout among oncology nurses. International Journal of Nursing Practice, 25, 1–11. https://doi. org/10.1111/ijn.12736 Yerlikaya, E. E. (2009). Üniversite öğrencilerinin mizah tarzları ile algıla- nan stres, kaygı ve depresyon düzeyleri arasındaki ilişkinin incelenmesi (Yayımlanmamış doktora tezi). Adana: Çukurova Üniversitesi. Yeşil, A., Ergün, Ü., Amasyalı, C., Er, F., Olgun, N. N., & Aker, A. T. (2010). Çalışanlar için yaşam kalitesi ölçeği Türkçe uyarlaması geçerlik ve güvenirlik çalışması. Nöropsikiyatri Arşivi, 47, 111–117. https://doi. org/10.4274/npa.5210 Yu, H., Jiang, A., & Shen, J. (2016). Prevalance and predictors of com- passion fatigue, burnout and compassion satisfaction among oncol- ogy nurses: A cross-sectional survey. International Journal of Nursing Studies, 57, 28–38. https://doi.org/10.1016/j.ijnur stu.2016.01.012 How to cite this article: Pehlivan T, Güner P. Effect of a compassion fatigue resiliency program on nurses’ professional quality of life, perceived stress, resilience: A randomized controlled trial. J Adv Nurs. 2020;76:3584–3596. https://doi. org/10.1111/jan.14568 https://doi.org/10.5388/aon.2017.17.2.116 https://doi.org/10.1111/jocn.13330 https://doi.org/10.1188/17.CJON.259-262 https://doi.org/10.4040/jkan.2016.46.2.260 https://doi.org/10.4040/jkan.2016.46.2.260 https://doi.org/10.20467/1091-5710.17.4.52 https://doi.org/10.1097/JTN.0000000000000284 https://doi.org/10.1080/13561820.2017.1309364 https://doi.org/10.1080/13561820.2017.1309364 https://doi.org/10.1188/13.ONF.180-187 https://doi.org/10.1188/13.ONF.180-187 https://doi.org/10.1097/NAQ.0b013e3182a2f9dd https://doi.org/10.1188/15.ONF.198-199 https://doi.org/10.1188/15.ONF.198-199 https://doi.org/10.4037/ajcc2015291 https://doi.org/10.4037/ajcc2015291 https://doi.org/10.1177/1049732305283930 https://doi.org/10.1016/j.explore.2012.10.002 https://doi.org/10.1188/13.ONF.E327-E336 https://doi.org/10.4103/2347-5625.135818 https://doi.org/10.1186/s12904-018-0346-9 https://doi.org/10.1186/s12904-018-0346-9 https://doi.org/10.1111/ijn.12736 https://doi.org/10.1111/ijn.12736 https://doi.org/10.4274/npa.5210 https://doi.org/10.4274/npa.5210 https://doi.org/10.1016/j.ijnurstu.2016.01.012 https://doi.org/10.1111/jan.14568 https://doi.org/10.1111/jan.14568 3596  |     PEHLIVAN ANd GÜNER The Journal of Advanced Nursing (JAN) is an international, peer-reviewed, scientific journal. JAN contributes to the advancement of evidence-based nursing, midwifery and health care by disseminating high quality research and scholarship of contemporary relevance and with potential to advance knowledge for practice, education, management or policy. JAN publishes research reviews, original research reports and methodological and theoretical papers. For further information, please visit JAN on the Wiley Online Library website: www.wileyonlinelibrary.com/journal/jan Reasons to publish your work in JAN: • High-impact forum: the world’s most cited nursing journal, with an Impact Factor of 1.998 – ranked 12/114 in the 2016 ISI Journal Citation Reports © (Nursing (Social Science)). • Most read nursing journal in the world: over 3 million articles downloaded online per year and accessible in over 10,000 libraries worldwide (including over 3,500 in developing countries with free or low cost access). • Fast and easy online submission: online submission at http://mc.manuscriptcentral.com/jan. • Positive publishing experience: rapid double-blind peer review with constructive feedback. • Rapid online publication in five weeks: average time from final manuscript arriving in production to online publication. • Online Open: the option to pay to make your article freely and openly accessible to non-subscribers upon publication on Wiley Online Library, as well as the option to deposit the article in your own or your funding agency’s preferred archive (e.g. PubMed). QOL of Home Health Care Patients at SciVerse ScienceDirect Asian Nursing Research 7 (2013) 53e60 Contents lists available Asian Nursing Research journal homepage: www.asian-nursingresearch.com Research Article Clinical Outcomes and Quality of Life of Home Health Care Patients Suk Jung Han, PhD, RN,1,* Hyun Kyung Kim, PhD, RN, 2 Judith Storfjell, PhD, RN, FAAN, 3 Mi Ja Kim, PhD, RN, FAAN 3 1 Department of Nursing, Sahmyook University, Seoul, South Korea 2 College of Nursing, Chonbuk Research Institute of Nursing Science, Chonbuk National University, Jeonju, South Korea 3 College of Nursing, University of Illinois at Chicago, Chicago, United States a r t i c l e i n f o Article history: Received 27 April 2012 Received in revised form 17 November 2012 Accepted 27 February 2013 Keywords: activities of daily living home care services quality of life * Correspondence to: Suk Jung Han, PhD, RN, Depa University, Hwarangro-815 Nowon-gu, Seoul 139-742 E-mail address: fountain@syu.ac.kr 1976-1317/$ e see front matter Copyright � 2013, Ko http://dx.doi.org/10.1016/j.anr.2013.03.002 s u m m a r y Purpose: This study aimed to evaluate the quality of life (QOL) in home health care patients according to change in health status outcomes between the start of care and discharge or 60 days, whichever came first. Methods: This is a prospective descriptive study. The convenience sample consisted of 100 home health care patients, who started receiving home health care services from a home health care agency in the United States. The World Health Organization Quality of Life Scale-Brief (WHOQOL-BREF) was used for measuring QOL; activities of daily living (ADLs) and instrumental ADLs were collected from the Outcome and Assessment Information Set data via Centers for Medicare and Medicaid Services-required home health agencies. Descriptive statistics, paired t tests, and multiple linear regressions were used for data analysis. Results: ADLs and instrumental ADLs of participants significantly improved between start of care and discharge or 60 days. Overall QOL, general health, and three of four QOL domains (physical, psychological, and environmental, but not social domain) were significantly improved at discharge or 60 days. Conclusion: Home health care nurses should maintain and improve the functional ability of patients, as this could improve the QOL of these patients. Copyright � 2013, Korean Society of Nursing Science. Published by Elsevier. All rights reserved. Introduction The home health care delivery system in the United States has expanded as the demand for the care of acute/chronic health problems increased, particularly among the growing elderly pop- ulation (Kirby & Lau, 2010). Medicare-certified home health agencies grew in number from 6,809 in 2001 to 10,422 in 2008 (Dey, Johnson, Pagerowski, Tanamor, & Ward, 2011) and provided care for more than 3 million Americans in 2010 (Centers for Medicare & Medicaid Services [CMS], 2010). Contemporary socio- political and economic forces have also influenced the home health care environment (Dieckmann, 2005). Following the downturn of Medicare home health care by the Balanced Budget Act in 1997, home health care began to recover under the home health pro- spective payment system (Murkofsky & Alston, 2009). Demand for home care services increased not only because of increasing elderly population, but also because of consumer preference and rtment of Nursing, Sahmyook , South Korea. rean Society of Nursing Science. P technological advances that allowed complex care to be delivered at home (Ellenbecker, Porell, Samia, Byleckie, & Milburn, 2008). Home health care services are available to all age groups, but 70.5% of such patients were elderly people aged 65 years or above (Caffrey, Sengupta, Moss, Harris-Kojetin, & Valverde, 2011; National Center for Health Statistics, 2005). In the United States, utilization of home health care services peaked in 1996 with 90.6 individuals per 10,000 of population, but it decreased to 48.7 patients per 10,000 in 2000 (National Center for Health Statistics). Determining objective outcomes of care became an important issue as home health care visits and expenditures grew (Shaughnessy et al., 1996). Hence, the CMS required home health agencies to submit Outcome and Assessment Information Set (OASIS) data for reimbursement. Thus, the CMS began to report publicly the OASIS outcome data for all home health agencies in the United States (CMS, 2003). OASIS is a tool that evaluates the outcomes of home health services (Shaughnessy & Crisler, 2005). It is a 79-item instrument developed to provide a standardized collection of outcomes data in the home health care setting (Shaughnessy et al., 2002). Its results can be used for outcome-based quality improvement, prospective ublished by Elsevier. All rights reserved. mailto:fountain@syu.ac.kr www.sciencedirect.com/science/journal/19761317 http://www.asian-nursingresearch.com http://dx.doi.org/10.1016/j.anr.2013.03.002 http://dx.doi.org/10.1016/j.anr.2013.03.002 http://dx.doi.org/10.1016/j.anr.2013.03.002 S.J. Han et al. / Asian Nursing Research 7 (2013) 53e6054 pay, and public reporting of quality data through the “Home Health Compare” initiative (CMS, 2011). OASIS-based quality performance has been reported by the CMS since 2003; it has shown how well home health agencies assisted their patients in regaining or maintaining their ability to function (CMS, 2011). Each agency’s success in achieving positive outcomes on designated OASIS mea- sures was compared to the agency’s previous performance and to that of other agencies (Keepnews, Capitman, & Rosati, 2004). Evaluation of OASIS data focuses on agency performance on specific patient outcomes, including changes in a patient’s health status between two or more time points. While OASIS provides basic data on the outcomes of home health services, it lacks a measure of quality of life (QOL) of care recipients. QOL has been used increasingly as an important parameter of health and well- being. QOL is defined as individuals’ perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards, and concerns (WHOQOL Group, 1994). In clinical practice and clinical trials, QOL indicators are used to evaluate treatment in terms of human costs and benefits. QOL has also been used to make de- cisions regarding allocation of health care services (Hadorn, 1991). Contemporary goals of treatment are expected to include reducing the severity of symptoms, improving functional status, and improving general QOL (Fletcher, Hunt, & Bulpitt, 1987). QOL research may help predict the course of disease, the process of recovery, the usefulness of therapeutic interventions, the need for specific services, or prognostic indicators of survival duration (DeVon & Ferrans, 2003; Montazeri, 2009). As the population ages and health care evolves in its emphasis from acute care to chronic care, the assessment of QOL will help our comprehensive under- standing of its relationship with patient outcomes of home health care service (Fortinsky & Madigan, 2004). Individuals’ subjective perceptions of physical health, psychological health, social func- tioning, and environment are called "subjective quality of life" and are independent determinants of wellness and disease burden in patients. There is general agreement that subjective QOL is a multi- factor-determined construct (Jung et al., 2012). Even when medical treatment may appear successful, home care patients may have poor psychosocial functioning, adjustment to illness, or QOL. Hence, inclusion of QOL measures could allow for more comprehensive evaluation of the efficacy of treatment and/or home health care (Buck, Jacoby, Massey, & Ford, 2000). Yet few studies have addressed how home health care influences patients’ QOL. This research aimed to fill the gap and compared QOL outcomes as well as functional outcomes of patients who received home health care service between start of care and discharge or 60 days, whichever came first. Client outcomes are defined as changes in health status resulting from health care activities or interventions (Urden, 2001). Improved client outcomes often result from the combined effect of personal resources and activities plus assistance from professional providers (Holzemer, 1992). The specific aims of this study were to (a) compare the clinical outcomes of home health care patients between the start of care and discharge or 60 days, whichever came first, (b) compare home health care patients’ QOL between start of care and at 60 days or at discharge and (c) identify the determinants of changes in quality of life in home health care patients. Methods Study design This is a prospective descriptive study using a structured questionnaire. Setting and samples A convenience sample of 110 patients who received home health care services from a home health care agency was recruited; the agency was affiliated with a hospital in a metropolitan city in the United States. The inclusion criteria were the following: Patients were (a) enrolled in a certified home health care agency, (b) were 45 years old and older, (c) had intact cognitive status and no mental disorder, and (d) had no difficulty in communication in English. The sample size was calculated by G*Power program using alpha at .05, power (1 e b) ¼ .95, and an effect size of .35. The estimated sample size was 70. Ethical considerations Institutional review board (protocol no. 2008-0445) approvals were received from the University of Illinois at Chicago and the hospital that had formal affiliation with the home health care agency. Patients who met the inclusion criteria were identified and approached by home health care nurses who were trained for the study. Those who agreed to participate completed a consent form, and arrangements were made for an appointment to complete the questionnaire. Measurements Clinical outcomes Patient clinical outcome was measured at start of care and at discharge or after 60 days of service, whichever came first. Clinical outcomes were assessed by using a total of 14 items from OASIS, which included 8 items in activities of daily living (ADLs) and 6 items in instrumental activities of daily living (IADLs). While the ADL represents health status as well as necessary ability for inde- pendent living, IADL reflects the level of instrumental indepen- dence. ADL items were grooming, ability to dress upper body, ability to dress lower body, bathing, ability to wash entire body, toileting, transferring, ambulation/locomotion, and feeding or eating. Items for IADL included planning and preparing light meals, transportation, laundry, housekeeping, shopping, and ability to use telephone. Within the OASIS, individual items assess different as- pects of functional performance. The individual items have different levels of scoring. For all ADLs and IADLs, a value of 0 in- dicates complete independence and is the best score possible. Ac- cording to Scharpf and Madigan (2010), using the corrected Likert approach puts all of the individual ADLs and IADLs on the same scale, ranging from 0 to 1. For ease of interpretation across all items, they were reversely coded (Table 1). The total clinical outcome score was then computed by sum- ming the individually adjusted items for a range in ADLs from 0 to 8, with 0 indicating total dependence in functional items as a group and 8 indicating complete independence, and in IADLs a range from 0 to 6, with 0 indicating total dependence in all items as a group and 6 indicating complete independence. Higher scores indicated higher clinical outcomes. To combine the 8 ADL items that were measured with different scales into a single index, each item was rescored on a scale of 0e1 and then recoded to reverse the direction of scoring. The same process was applied to the 6 IADL items. Recoding was done to reflect level of independence, rather than dependence; this enhanced ease of interpretation of the results: the magnitude of improvement rather than decline was evaluated (Keepnews et al., 2004). "Improvement" meant improved status of ADL and IADL from start of care to discharge or 60 days, and "decline" and "un- changed" meant, respectively, decreased status or no changes in ADL and IADL between the two time points. Table 2 General Characteristics of Participants (N ¼ 100) Characteristics Categories n (%) M (SD) Age (yr) �64 37 (37.0) 69.90 (11.92) 65e74 26 (26.0) 75e84 26 (26.0) �85 11 (11.0) Gender Male 37 (37.0) Female 63 (63.0) Race/ethnicity White 80 (80.0) Asian 8 (8.0) Black or African American 10 (10.0) Hispanic or Latino 1 (1.0) Native Hawaiian or Pacific Islander 1 (1.0) Marital status Married 33 (33.0) Single 29 (29.0) Widowed 25 (25.0) Divorced 13 (13.0) Education �5th grade 4 (4.0) Junior high school 4 (4.0) High school 43 (43.0) College 32 (32.0) >College 17 (17.0)

Annual income < $20,000 40 (40.0) $20,000e30,000 10 (10.0) $30,000e40,000 9 (9.0) $40,000e50,000 9 (9.0) >$50,000 21 (21.0)
Unknown 11 (11.0)

Current
residence

Patient-owned or
rented residence

83 (83.0)

Family member’s
residence

14 (14.0)

Board & care or
assisted living
facility

2 (2.0)

Others 1 (1.0)
Living

arrangementa
Alone 38 (42.7)
With spouse or
significant other

27 (30.3)

With other family
member

22 (24.7)

With a friend 2 (2.2)
Payment source Medicare 33 (33.0)

Medicaid 9 (9.0)
Private insurance 28 (28.0)
Medicare þ private

insurance
27 (27.0)

Others 3 (3.0)
Primary caregiver No 21 (21.0)

Spouse or significant
other

32 (32.0)

Daughter or son 27 (27.0)
Other family member 13 (13.0)
Friend, neighbor,
community or church
member

3 (3.0)

Paid help 4 (4.0)
ICD-9 code Health services for specific

procedures & after care
42 (42.0)

Disease of the
respiratory system

14 (14.0)

Disease of the
circulatory system

13 (13.0)

Injury and poisons (e.g.,
fracture, open wound)

11 (11.0)

Disease of the skin 6 (6.0)
Symptom, signs and
ill-defined condition

6 (6.0)

Table 1 OASIS Measures Used in Functional Status Index

OASIS item Range Conversion to ADL 8 and IADL 6 items

0 1 2 3 4 5

ADL items
Grooming 0e3 1 .67 .33 0
Dress upper body 0e3 1 .67 .33 0
Dress lower body 0e3 1 .67 .33 0
Bathing 0e5 1 .80 .60 .40 .20 0
Toileting 0e4 1 .75 .50 .25 0
Transferring 0e5 1 .80 .60 .40 .20 0
Ambulation 0e5 1 .80 .60 .40 .20 0
Feeding or eating 0e5 1 .80 .60 .40 .20 0

IADL items
Plan/prepare light meals 0e2 1 .50 0
Transportation 0e2 1 .50 0
Laundry 0e2 1 .50 0
Housekeeping 0e4 1 .75 .50 .25 0
Shopping 0e3 1 .67 .33 0
Telephone 0e5 1 .80 .60 .40 .20 0

Note. OASIS ¼ outcome and assessment information set; ADL ¼ activities of daily
living; IADL ¼ instrumental activities of daily living.

S.J. Han et al. / Asian Nursing Research 7 (2013) 53e60 55
Internal consistency, Cronbach’s coefficient alphas were .86 in
ADL and .75 in IADL in the present study. Only one study that re-
ported the criterion-related validity of the scales was found,
showing a correlation of .44e.69 in the ADL and .20e.68 in the IADL
(Tullai-McGuinness, Madigan, & Fortinsky, 2009).

QOL
The World Health Organization Quality of Life Scale-Brief

(WHOQOL-BREF) questionnaire (Bonomi & Patrick, 1997;
WHOQoL Group, 1995) was used to measure perceived QOL. It
consists of 2 global items (overall QOL and general health) and 24
items in the domains of physical, psychological, social relationship,
and environmental. The physical domain covers pain, energy, and
medication needs (7 items). The psychological domain explores
feelings about the meaning of life, capacity of concentration, physical
appearance, and feelings of desperation (6 items). The social re-
lationships domain is concerned with friend support and sexual
satisfaction (3 items). The environmental domain covers perceived
security in daily life, individual satisfaction about transport, and
personal impressions about health services (8 items). The score of
each domain is found by multiplying the calculated mean value of
the items belonging to the component and thus ranges from 4 to 20
(Skevington, Lotfy, & O’Connell, 2004).

Internal consistency, Cronbach’s alphas were between .66 and
.84 when it was developed (WHO,1998). In the present study, these
were between .73 and .91 (.91 in total quality of life, .81 in physical
domain, .76 in psychological domain, .73 in social relationship
domain, .78 in environmental domain). The validity and reliability
were tested in the study of Skevington et al. (2004) with a sample
size of 11,830 from 23 countries.

General characteristics
General characteristics included items such as age, gender, race/

ethnicity, marital status, education, annual income status, current
residence, living arrangement, primary caregiver, patient’s disease
classification according to the International Classification of
Diseases-9 code, payment source, duration of service (from start of
care to discharge or 60 days ), and number of registered nurse visits.
Neoplasms 4 (4.0)
Disease of the nervous
system

2 (2.0)

(continued on next page)
Data collection

Data were collected using a questionnaire between November
2008 and December 2009. Trained home health care nurses visited

Table 2 (continued )

Characteristics Categories n (%) M (SD)

Service duration
(day)

�20 38 (38.0) 30.35 (18.00)
21e30 25 (25.0)
31e40 15 (15.0)
41e50 9 (9.0)
51e60 13 (13.0)

No. of RN visits �5 28 (28.0) 8.54 (5.62)
6e7 29 (20.0)
8e9 18 (27.0)
�10 25 (25.0)

Note. ICD-9 ¼ international classification of diseases; RN ¼ registered nurse.
a Missing value excluded.

S.J. Han et al. / Asian Nursing Research 7 (2013) 53e6056
home health care patients and asked if they would be willing to
participate in a research project. If they agreed, home health nurses
explained about the study purpose and procedure and asked them
if they had any questions, and then obtained written consents.
Face-to-face interviews followed, using a structured questionnaire
that included a demographic profile. The questionnaires were
completed by home health nurses based on the interviews. ADL and
IADL were measured by home health nurses as a part of OASIS data
and QOL was self-reported. Each interview took 30e45 minutes.

Duration of home health care services prescribed by physician in
the United States is 60 days. The the order needs to be renewed
afterwards. In addition, OASIS data in the healthcare agencies are
collected as start of care, and at discharge or 60-day follow up.
Therefore the questionnaires were collected at the start of care and
at discharge or 60 days, whichever came first.

At the end of each interview, US$5 was given to patients for their
participation. Among the 110 participants, 1 was admitted to a
nursing home; 1 passed away; 8 patients did not finish the ques-
tionnaire at discharge or 60 days. A total of 100 patients completed
in this study.

Data analysis

Data were analyzed by SPSS version 17.0 (SPSS Inc., Chicago, IL,
USA). Descriptive statistics were used to summarize all data. Paired
t test was used to compare data between admission and at
discharge or 60 days. McNemar’s test was used to compare
Table 3 Comparison of ADL and IADL between SOC and Discharge or 60 Days (N ¼ 100)

Variables SOC Discharge Change t p

M (SD) M (SD) M (SD)

ADL Grooming 0.74 (0.18) 0.93 (0.16) 0.19 (0.21) 9.06 <. Dress upper body 0.70 (0.21) 0.92 (0.18) 0.22 (0.19) 11.33 <. Dress lower body 0.56 (0.28) 0.86 (0.23) 0.30 (0.25) 12.10 <. Bathing 0.50 (0.29) 0.74 (0.26) 0.24 (0.27) 8.74 <. Toileting 0.91 (0.15) 0.98 (0.12) 0.07 (0.15) 4.38 <. Transferring 0.83 (0.10) 0.89 (0.11) 0.06 (0.10) 5.67 <. Ambulation 0.81 (0.12) 0.86 (0.10) 0.05 (0.10) 5.22 <. Feeding or eating 0.88 (0.12) 0.97 (0.10) 0.09 (0.10) 9.00 <. Total ADL 5.93 (1.11) 7.14 (0.92) 1.21 (0.91) 13.33 <. IADL Plan/prepare light meals 0.50 (0.31) 0.86 (0.28) 0.35 (0.30) 11.77 <. Transportation 0.50 (0.10) 0.59 (0.19) 0.09 (0.19) 4.50 <. Laundry 0.07 (0.17) 0.70 (0.37) 0.24 (0.33) 7.29 <. Housekeeping 0.27 (0.34) 0.56 (0.37) 0.29 (0.36) 8.23 <. Shopping 0.28 (0.20) 0.52 (0.28) 0.24 (0.24) 10.27 <. Telephone 0.96 (0.14) 0.97 (0.12) 0.02 (0.11) 1.69 . Total IADL 2.57 (0.88) 3.80 (1.21) 1.23 (0.99) 12.38 <. Note. ADL ¼ activities of daily living; IADL ¼ instrumental activities of daily living; SOC a p is calculated by paired t test. b Exact Chi-square test for equal proportion among improvement, decline, and uncha percentage of participants reporting good QOL and bad QOL be- tween start of care and discharge or 60 days. Chi-square test was used to compare proportion differences among improvement, decline, and unchanged in ADL and IADL between the two time points. Multiple linear regression was used to identify factors which affected the differences in QOL. Results General characteristics of participants Participant characteristics are shown in Table 2. As can be seen, the mean age of the 100 participants was 69.9 years, and 63% of them were over 65 years old. The sample was predominantly fe- male (63%), White (80%), had at least a high school education (92%), and lived in their own homes or rented residences (83%). Among participants, 33% were married and 29% single; 43% lived alone, and 30% lived with a spouse or significant other. Home health fees were reimbursed by Medicare (33%), private insurance (28%), or both Medicare and private insurance (27%). Primary caregivers were spouses or significant others for 32%. According to the International Classification of Diseases-9 codes, 42% of participants were cate- gorized as "health services for specific procedures and after care". Home care service was received for 30 days or less by 63%, and the mean frequency of registered nurse visits was 8.5 days, ranging from 3 to 38 days. Clinical outcomes between the start of care and discharge or 60 days Clinical outcomes of participants between the start of care and discharge or 60 days, as measured by OASIS items, are presented in Table 3. The total mean score of ADL at the start of care and at discharge or 60 days was 5.93 and 7.14, respectively, showing a significant increase of 1.21 (t ¼ 13.33, p < .001). The total mean score of IADL at the start of care and discharge or 60 days was 2.57 and 3.80, respectively, showing a significant increase of 1.23 (t ¼ 12.38, p < .001). As measured by the sum of the 8 ADL items, 90% of participants showed improvement in total ADL, 3% showed a decline, and 7% remained unchanged (c2 ¼ 144.74, p < .001). In terms of individual a Unit change c2b p Improvement (%) Decline (%) Unchanged (%) 1 unit �2 units 1 unit �2 units 001 54 2 1 1 42 47.12 <.001 001 58 4 1 0 37 56.00 <.001 001 47 21 0 0 32 12.96 <.001 001 30 33 0 2 35 55.94 <.001 001 27 1 0 1 71 74.78 <.001 001 30 0 2 0 68 65.84 <.001 001 24 1 1 0 74 83.06 <.001 001 45 0 0 0 55 1.00 .317 001 90% 3% 7% 144.74 <.001 001 59 6 1 0 34 61.46 <.001 001 17 0 0 0 83 43.56 <.001 001 33 8 1 0 58 51.38 <.001 001 16 35 2 0 47 44.42 <.001 001 46 13 0 0 41 3.24 .072 095 3 4 1 1 91 150.02 <.001 001 90% 1% 9% 145.46 <.001 ¼ start of care. nged. S.J. Han et al. / Asian Nursing Research 7 (2013) 53e60 57 ADL items, participants showed the most improvement in dress lower body (68%), followed by bathing (63%) and dress upper body (62%). Unit change means the change of score between start of care and discharge. If the score at start of care was 1 and then the score changes to 0 at discharge, that is called 1 unit change. ADL items that showed marked improvement by more than 2 units included bathing (33%) and dress lower body (21%). On the other hand, 0e2% of participants showed a decline, and 32e74% remained unchanged for these same items. The most unchanged item was ambulation (74%), followed by toileting (71%) and transferring (68%). As measured by the sum of the 6 IADL items, 90% of participants showed improvement, 1% declined, and 9% remained unchanged (c2 ¼ 145.46, p < .001). When analyzed by individual IADL items, the percentages of improved participants varied from 7% to 65%. Plan/prepare light meals (65%) was the most improved item, fol- lowed by shopping (59%) and housekeeping (51%). The most markedly improved item by more than 2 units was housekeeping (35%). On the other hand, 0e2% of participants showed a decline in housekeeping. For each IADL item, 34e91% of participants remained unchanged. The most unchanged item was ability to use telephone (91%), followed by transportation (83%) and laundry (58%). QOL between the start of care and discharge or 60 days A comparison of QOL between the start of care and discharge or 60 days is shown in Table 4. Overall QOL was significantly improved at discharge or 60 days compared to that at start of care, showing a mean increase of 0.37 (t ¼ 3.46, p ¼ .001). Sixty percent of partic- ipants at start of care and 85% at discharge or 60 days reported good QOL, while the difference between the two time points was sig- nificant by McNemar’s test (p < .001). General health at discharge or 60 days was significantly improved compared to that at the start of care, showing a mean increase of 0.85 (t ¼ 7.51, p < .001). Thirty percent of participants at the start of care and 74% at discharge or 60 days reported being satisfied with their general health (p < .001). Mean domain scores ranged from 12.09 (physical domain) to 16.95 (social domain) at start of care and from 15.17 (physical domain) to 17.33 (social domain) at discharge or 60 days. The biggest change was in the physical domain (M ¼ 3.08, SD ¼ 2.73). All QOL domains except the social domain were significantly improved at discharge or 60 days compared to the level at the start of care, including physical (t ¼ 10.58, p < .001), psychosocial (t ¼ 5.18, p < .001), and environmental (t ¼ 6.58, p < .001). Spe- cifically, the physical domain, which showed the lowest score at start of care (M ¼ 12.09, SD ¼ 2.41), was most improved at discharge Table 4 Comparison of QOL between SOC and Discharge or 60 Days Variables (range) SOC Discharge Change t pa M (SD) M (SD) M (SD) Overall QOL (1e5) 3.70 (0.98) 4.06 (0.73) 0.37 (1.04) 3.46 .001 Good n (%) 60 (60) 85 (85) <.001b Not good n (%) 40 (40) 15 (15) General health (1e5) 2.98 (1.04) 3.83 (0.90) 0.85 (1.13) 7.51 <.001 Satisfied n (%) 30 (30) 74 (74) <.001b Unsatisfied n (%) 70 (70) 26 (26) Physical (4e20) 12.09 (2.41) 15.17 (2.31) 3.08 (2.73) 10.58 <.001 Psychological (4e20) 14.80 (2.28) 16.13 (2.19) 1.33 (2.48) 5.18 <.001 Social (4e20) 16.95 (2.62) 17.33 (2.34) 0.38 (2.21) 1.73 .087 Environmental (4e20) 15.49 (2.01) 16.80 (1.94) 1.31 (1.88) 6.58 <.001 Note. QOL ¼ quality of life; SOC ¼ start of care. a p calculated by paired t test. b p calculated by McNemar test. or 60 days (M ¼ 15.17, SD ¼ 2.31), whereas the social domain, which showed the highest score at start of care (M ¼ 16.95, SD ¼ 2.62), was least improved at discharge or 60 days (M ¼ 17.23, SD ¼ 2.34). Determinants of change in QOL Multiple linear regressions were used to identify determinants of QOL change, and the results are shown in Table 5. Before using multiple linear regression, QOL change according to general char- acteristics, ADL, and IADL were tested using t test, analysis of variance, or correlation analysis. Only variables statistically signif- icant were included as independent variables for multiple linear regression. First, correlation between variables, tolerance, and variance inflation factor were tested to identify multicollinearity. The vari- ance inflation factor ranged from 1.201 to 4.007, indicating no multicollinearity. Independent variables included age, service duration, number of registered nurse visits, ADL at start of care, IADL at start of care, ADL change, and IADL change; dependent variables included the four domains of QOL. When analyzed by domains of QOL, the regression models of the domains physical (F ¼ 3.66, p ¼ .002), psychological (F ¼ 2.58, p ¼ .019), social (F ¼ 2.27, p ¼ .035) and environmental (F ¼ 5.22, p < .001) QOL were statistically significant. Physical domain was significantly explained by age, ADL at start of care, and ADL change (adjusted R2 ¼ .168, p ¼ .002). Psychological domain was signifi- cantly explained by ADL at start of care, ADL change, IADL at start of care, and IADL change (adjusted R2 ¼ .107, p ¼ .019). Social domain was significantly explained by the number of registered nurse visits only (adjusted R2 ¼ .083, p ¼ .035). Service duration, age, ADL at start of care, ADL change, and IADL change were statistically significantly explained by the change in environmental QOL (adjusted R2 ¼ .251, p < .001). Discussion The effects of home health care service on clinical outcomes, including ADL and IADL, and on changes in QOL were examined in 100 home care patients in one U.S. home care agency. Functional status is important because it is necessary for independent living and QOL (Scharpf & Madigan, 2010; Shaughnessy et al., 2002). Most recently, Madigan et al. (2012) reiterated the importance of func- tional capacity as a key factor in maintaining the ability of older people to live independently and safely at home and as a key focus area for home health care. In this study, both clinical outcomes (in terms of ADL and IADL) and QOL were significantly improved at discharge or 60 days compared with the levels at the start of care. Clinical outcomes, ADL, and IADL The ADL score included capacity for daily self-care, which is essential for ensuring independent living and contributes impor- tantly to overall QOL (Drewnowski & Evans, 2001). Maintaining daily functions without assistance may be the most salient outcome variable. Seven million Americans aged more than 65 years depend on others for help with some basic tasks of daily living (Ory & Cox, 1994). While the ADL data represent health status as well as necessary ability for independent living, IADL reflects the level of instrumental independence. In this study, 90% of participants improved in both ADL and IADL. This finding compares favorably with that of Keepnews et al. (2004), who reported improvement of both ADL and IADL in 78% of 1,051 home care patients who received home health care for less than 60 days. Hadley, Rabin, Epstein, Stein, and Rimes (2000) examined functional status of patients at the time of discharge Table 5 Factors Influencing the Relationship of Functional Outcome and QOL between SOC and Discharge or 60 Days Variables QOL domains Physical Psychological Social Environmental Ba bb t p Ba bb t p Ba bb t p Ba bb t p (Constant) 11.17 2.02 .047 e6.33 2.12 .037 e3.53 1.35 .182 7.94 3.63 <.001 Service duration e0.02 e0.07 e0.60 .553 e0.01 e0.08 e0.71 .477 0.01 0.07 0.63 .528 0.03 0.27 2.48 .015 No. of RN visit e0.04 e0.05 e0.45 .653 0.06 0.14 1.18 .240 e0.10 e0.25 e2.23 .028 e0.04 e0.10 e0.92 .360 Age 0.09 0.22 2.06 .042 0.04 0.20 1.80 .075 0.04 0.19 1.81 .073 0.05 0.32 3.21 .002 ADL at SOC 2.05 0.49 2.53 .013 1.24 0.54 2.84 .006 0.43 0.22 1.13 .262 0.91 0.52 2.92 .005 ADL change 2.16 0.42 2.86 .005 1.21 0.44 2.94 .004 0.51 0.21 1.43 .155 1.28 0.61 4.39 <.001 IADL at SOC 0.77 0.15 0.94 .350 1.17 0.42 2.70 .008 0.40 0.16 1.02 .311 0.46 0.22 1.53 .130 IADL change 1.03 0.22 1.74 .086 0.91 0.34 2.71 .008 0.15 0.07 0.55 .585 0.49 0.27 2.21 .030 Adjusted R2 ¼ .168 Adjusted R2 ¼ .107 Adjusted R2 ¼ .083 Adjusted R2 ¼ .251 F ¼ 3.66 F ¼ 2.58 F ¼ 2.27 F ¼ 5.22 p ¼ .002 p ¼ .019 p ¼ .035 p < .001 Note. QOL ¼ quality of life; SOC ¼ start of care; RN ¼ registered nurse. a B is the unstandardized regression coefficient. b b is the standardized regression coefficient. S.J. Han et al. / Asian Nursing Research 7 (2013) 53e6058 from the hospital and 6 months later. They reported that functional status of patients who received home health care showed more improvement than that of nonusers of the home care service. Scharpf and Madigan (2010) compared ADL in OASIS of patients with heart failure who received home health care at start of care and at discharge, and found that 86% experienced improvement or stayed the same. In that study, each item of ADL improved from 0.03 to 0.17 compared to the improvement from 0.05 to 0.30 in this study. The most improved item was dress lower body followed by bathing, the same finding as in Scharpf and Madigan’s research on functional status outcome measures in home health care patients with heart failure. Hadley et al. (2000) analyzed the effect of posthospitalization home health care use on the change in functional status for a sample of 2,127 (over 65 years of age) Medicare beneficiaries who participated in Medicare’s Current Beneficiary Survey for 6 months after hospital discharge. Home health care users experienced greater improvement in functional status than nonusers, as measured by the change in a continuous scale based on the number and mix of ADL and IADL before and after hospitalization. The estimated improvement in functional status could be as large as 13% for a 10% increase in home health care use. From a clinical perspective, it may be advantageous to use the individual ADL change scores, particularly the bathing score because bathing is a complex task, requiring multiple kinds of movements (transfer, use of upper and lower limbs) and may be a proxy representation of how well these patients can manage their self-care (Scharpf & Madigan, 2010). Studies about the effectiveness of home health care on stabi- lizing or improving patients’ functional status are limited, and the results were mixed. The conflicting findings may be due, in part, to the numerous chronic health problems experienced by home health care patients. In addition, home health care patients usually experience a downward trajectory of these conditions, requiring home health care goals aimed at slowing the progression of disease and minimizing symptoms rather than improving them. QOL and functional outcomes Subjects showed higher scores in most QOL items. Using 12.0 as the scale midpoint where QOL was judged to be neither good nor poor, the means indicated that QOL was above average. Skevington et al. (2004) analyzed QOL assessment as measured by WHOQOL- BREF from a survey of adults performed in 23 countries (n ¼ 11,830). The mean score in the United States for each domain of QOL ranged from 11.7 (social domain) to 15.5 (physical domain). The mean score of the physical domain at start of care in this study was lower, but the meanscores of other domains were higher compared to those in Skevington et al. (2004). Particularly, mean scores of all domains at discharge or 60 days in this study were higher than start of care. Generally, the higher QOL found in this study might be explained by differences in participants of the two studies, such as health status and income. In our sample, 80% were White, and only 9% got Medicaid service, whereas those in Skevington et al.’s study were sampled from the general population in hospitals, rehabilitation centers, and primary care settings with respect to quotas of important sociodemographic variables. The results of this study showed that QOL was significantly improved at discharge or 60 days compared to the QOL at start of care. When analyzed by domains, social domain was not signifi- cantly improved, and this might be related to considerably high scores of social QOL at start of care. Social domain is not affected by age, marital status, and edu- cation in rural areas. On the other hand, the presence of chronic disease and dependency in daily activities and lifestyle affect the social domain. In those with disease and the bedridden, social domain scores were the lowest (Arslantas, Unsal, Metintas, Koc, & Arslantas, 2009). Among ADLs, for the item of ambulation/loco- motion which is related to dependence, at start of care, only 1% of participants in this study were "chairfast, unable to ambulate but able to wheel self independently," 14% were “able to walk only with the supervision or assistance of another person at all times,” 65% “[required] use of a device to walk alone,” and 20% were “able to independently walk.” At discharge or 60 days, all were independent except for 2% who were “able to walk only with the supervision or assistance of another person at all times.” Although social QOL improved after home health care service, relatively high indepen- dence at start of care might have prevented the results from showing a statistically significant improvement. Naylor et al. (2004) examined the effectiveness of a 3-month Advanced Practice Nurse (APN)-directed discharge planning and home follow-up protocol (transitional care intervention) in elders with heart failure. The intervention group reported greater overall QOL at 12 weeks (p < .05), and in the physical dimension at 2 weeks (p < .01) and 12 weeks (p < .05). However, statistically significant group differences in functional status did not emerge, although less dependency was, on average, observed. Alexy, Benjamin-Coleman, and Brown (2001) did not find any significant changes in QOL be- tween start of care and discharge in studying functional status and QOL of Medicare home health clients at admission to home care and 30 days after admission. Difficulty in collecting both baseline and 30-day post home care admission data on each individual in a S.J. Han et al. / Asian Nursing Research 7 (2013) 53e60 59 timely fashion resulted in a very small sample (n ¼ 17), which was the likely reason for the nonsignificant findings. Helvik, Engedal, and Selbaek (2010) explored factors that affected QOL using the WHOQOL-BREF in older patients (M ¼ 82.8 years) who were admitted to the hospital. Their QOL was lower in all domains (physical domain at 12.6, environmental domain at 14.9) compared to that of this study. Their findings differed from our study, possibly due to their sample’s older age and severity of problems, as indicated by their admission to the hospital rather than to the nursing home. The multiple regression models of QOL do- mains and independent health-related variables were adjusted for each other. Three of four QOL domains (physical, psychological, and environmental, but not the social domain) were associated with ADL. Poor ADL was associated with a poorer QOL because a worse score in the ADL scale is an indicator of worse physical health. Poor physical health is known to influence QOL negatively (Helvik et al.). When Jeon and Choi (2010) investigated factors that influenced the health-related QOL of young-old men, old-old men, and oldest- old men in vulnerable age who received home care, they found correlations between IADL and health-related QOL of young-old (65e74 yr) (r ¼ .302, p < .05), old-old (75e84 yr) (r ¼ 315, p < .05), and oldest-old (85 or above) (r ¼ .293, p < .05). Also, IADL was one of the predictors in explaining the level of health-related QOL among vulnerable old men (who is Basic Livelihood Security and received home care from public health center). Tseng and Wang (2001) explored subjectively perceived QOL as measured by the QOL IndexeNursing Home Version and related factors of elderly nursing home residents. ADL (r ¼ .491, p < .05) had a significantly positive relationship with QOL, and ADL was one of the important predictors of QOL. Since the functional dependence level of patients and their ability to execute the ADL are meaningful to their QOL, these will have a direct effect on QOL. The relationships among the three di- mensions of QOL, need, and health behaviors were examined by Baernholdt, Hinton, Yan, Rose, and Mattos (2011) in a nationally representative sample of adults aged 65 years and older from the National Health and Nutrition Examination Survey (2005e2006). In bivariate analysis, they found that the need variable, ADL function, memory problems, and depression were associated with all three QOL dimensions, including health-related QOL, social functioning, and emotional well-being. However, only ADL was associated with all three dimensions of QOL in their full models, suggesting the importance of ADL to QOL. ADL and physical environment showed significant relationships with QOL in a study of community-based older adults in Canada (Low & Molzahn, 2007). Level of dependency in ADL and level of help received can affect the overall QOL (Hellstrom, Perssion, & Hallberg, 2004). As early as 1999, the Agency for Healthcare Research and Quality pointed out the importance of focusing on functional outcomes research. The OASIS tool offered an opportunity to incorporate standardized outcome data not previously available to home health care researchers (Keepnews et al., 2004). The CMS, administrator and payer of the Medicare program for aged and disabled Americans, has provided home health care agencies with several types of reports based on the OASIS. Internal agency re- ports included the number of patients whose conditions improved versus those who declined or stayed the same, at the individual ADL item level. There are also public reports, available on the "Home Health Compare" website. These provide agency-level in- formation on the percentage of patients who improved in specific ADL items. While helpful for targeting specific ADL items, a composite score may also be beneficial in agency quality improvement programs, as it would identify trends in patient populations. For instance, higher levels of functional impairments at discharge from home health care may require addressing clin- ical care differently during and after home health care (Scharpf & Madigan, 2010). Nurses recognized the importance of functional status as a pa- tient clinical/health outcome and as an important measure of quality of nursing care over two decades ago (American Nurses Association, 1992). Functional status was often measured by inde- pendence in ADLs and IADLs (Roberts, 1999). A meaningful differ- ence in ADL and IADL for home health care would guide policy and practice decisions for nurses as to what level of change is possible and attainable. Limitations This study explored the effect of home health care on the clinical outcomes and QOL. The study may have excluded other variables that could contribute to these variables. Future studies are needed to demonstrate the benefits of providing home health care with bigger sample sizes, a comparison group, and more comprehensive measures on QOL. Although ADL and IADL items in OASIS have been widely used at the home healthcare agencies in the United States, little evidence for the validity and relatively low criterion-related validity reported in a previous study (Tullai-McGuinness & Madigan, 2009) might limit the interpretation of the results in this study. Because the endpoint of measurement was defined in two ways (at discharge or 60 days), the time difference from start of care may not be the same for each subject. Conclusion Home care services provided by registered nurses using the items of OASIS have shown improved clinical outcomes and QOL after 60 days of home health care. Major improvements in ADL were dressing lower body and bathing and in IADL were preparing light meals, shopping, and housekeeping. ADL and IADL were important predictors of QOL in three of the four QOL domains (physical, psychological, and environmental, but not the social domain). Home health care nurses should focus on improving the functional ability of patients, as they play a key role in making a difference in the lives of these patients who stay at home and receive essential home care. Conflict of interest The authors declare no conflict of interest. This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Acknowledgments We thank Pat Koepp and Sharmila Rao for their support, and Suzanne Dunne, LeeAnn Wittenstrom, Conchita Penaranda, and Sun Ok Jang who helped collect data. We acknowledge Dr. Cathy Christenson for her guidance, Ji Sung Lee for statistical consultation, and Kevin Grandfield, Publication Manager for the University of Illinois at Chicago Department of Biobehavioral Health Science, for editorial assistance. References Alexy, B., Benjamin-Coleman, R., & Brown, S. (2001). Home healthcare and client outcomes. Home Healthcare Nurse, 19(4), 233e239. http://dx.doi.org/10.1097/ 00004045-200104000-00012 American Nurses Association. (1992). Scope and standards of home health nursing practice. Silver Spring, MD: Author. http://dx.doi.org/10.1097/00004045-200104000-00012 http://dx.doi.org/10.1097/00004045-200104000-00012 S.J. Han et al. / Asian Nursing Research 7 (2013) 53e6060 Arslantas, D., Unsal, A., Metintas, S., Koc, F., & Arslantas, A. (2009). Life quality and daily life activities of elderly people in rural areas, Eskisehir (Turkey). Archives of Gerontology and Geriatrics, 48(2), 127e131. http://dx.doi.org/10.1016/j.archger. 2007.11.005 Baernholdt, M., Hinton, I., Yan, G., Rose, K., & Mattos, M. (2011). Factors associated with quality of life in older adults in the United States. Quality of Life Research, 21(3), 527e534. http://dx.doi.org/10.1007/s11136-011-9954-z Bonomi, A. E., & Patrick, D. L. (1997). Users manual and interpretation guide of the United States version of the World Health Organization Quality of Life (WHOQOL) instrument. Seattle, WA: U.S. WHO Quality of Life Center. Buck, D., Jacoby, A., Massey, A., & Ford, G. (2000). Evaluation of measures used to assess quality of life after stroke. American Heart Association, 31(8), 2004e2010. http://dx.doi.org/10.1161/01.STR.31.8.2004 Caffrey, C., Sengupta, M., Moss, A., Harris-Kojetin, L., & Valverde, R. (2011). Home health care and discharged hospice care patients: United States, 2000 and 2007. National Health Statistics Reports, 38, 1e28. Centers for Medicare and Medicaid Services. (2003). Home health quality initiative overview. Retrieved from. http://www.cms.hhs.gov/HomeHealthQualityInitis/ Downloads/HHQIOverview/pdf Centers for Medicare and Medicaid Services. (2010). 2010 data compendium. Retrieved from. https://www.cms.gov/DataCompendium/14_2010_Data_Compendium.asp Centers for Medicare and Medicaid Services. (2011). Home health quality initiatives. Retrieved from. http://www.cms.gov/HomeHealthQualityInits/ DeVon, H. A., & Ferrans, C. E. (2003). The psychometric properties of four quality of life instruments used in cardiovascular populations. Journal of Cardiopulmonary Rehabilitation, 23(2), 122e138. Dey, J. G., Johnson, M., Pagerowski, W., Tanamor, M., & Ward, A. (2011). Home health study report. Washington, DC: L&M Policy Research. Retrieved from. http:// www.lmpolicyresearch.com/sites/default/files/hhpps_literaturereview Dieckmann, J. L. (2005). Home health administration: an overview. Handbook of home health care administration (4th ed.). Sudbury, MA: Jones and Bartlett. Drewnowski, A., & Evans, W. J. (2001). Nutrition, physical activity, and quality of life in older adults: summary. Journal of Gerontology, 56A(Special Issue II), 89e94. Ellenbecker, C. H., Porell, F. W., Samia, L., Byleckie, J. J., & Milburn, M. (2008). Pre- dictors of home health nurse retention. Journal of Nursing Scholarship, 40(2), 151e160. http://dx.doi.org/10.1111/j.1547-5069.2008.00220.x Fletcher, A. E., Hunt, B. M., & Bulpitt, C. J. (1987). Evaluation of quality of life in clinical trials of cardiovascular disease. Journal of Chronic Disease, 40(6), 557e566. http://dx.doi.org/10.1016/0021-9681(87)90014-2 Fortinsky, R. H., & Madigan, E. A. (2004). Data, information, and quality indicators for home healthcare: rapid implementation, what’s next? Journal for Healthcare Quality, 26(3), 44e51. http://dx.doi.org/10.1111/j.1945-1474.2004.tb00495.x Hadley, J., Rabin, D., Epstein, A., Stein, S., & Rimes, C. (2000). Posthospitalization home healthcare use and changes in functional status in a Medicare population. Medical Care, 38(5), 494e507. Hadorn, D. (1991). The Oregon priority-setting exercise: quality of life and public policy. Hastings Center Report, 21(3), 11e16. http://dx.doi.org/10.2307/3563329 Hellstrom, Y., Perssion, G., & Hallberg, I. R. (2004). Quality of life and symptoms among older people living at home. Journal of Advanced Nursing, 48(6), 584e593. http://dx.doi.org/10.1111/j.1365-2648.2004.03247.x Helvik, A. S., Engedal, K., & Selbaek, G. (2010). The quality of life and factors asso- ciated with it in the medically hospitalized elderly. Aging & Mental Health, 14(7), 861e869. http://dx.doi.org/10.1080/13607861003801003 Holzemer, W. I. (1992). Nursing effectiveness research and patient outcomes. A challenge for the second HIV/AIDS decade. Critical Care Nursing Clinics of North America, 4(3), 429e435. Jeon, E. Y., & Choi, Y. H. (2010). Factors affecting the health-related quality of life according to age in vulnerable aged man. Journal of Korean Academic Nursing, 40(3), 400e410. http://dx.doi.org/10.4040/jkan.2010.40.3.400 Jung, Y. E., Seo, H., Song, H. R., Woo, Y. S., Yim, H., Sung, H., et al. (2012). Factors associated with subjective quality of life in Korean patients with depressive disorders: the CRESCEND study. Quality of Life Research, 21(6), 967e974. http:// dx.doi.org/10.1007/s11136-011-0006-5 Keepnews, D., Capitman, J., & Rosati, R. (2004). Measuring patient-level clinical outcomes of home health care. Journal of Nursing Scholarship, 36(1), 79e85. http://dx.doi.org/10.1111/j.1547-5069.2004.04017.x Kirby, J. B., & Lau, D. T. (2010). Community and individual race/ethnicity and home health care use among elderly persons in the United States. Health Services Research, 45(5), 1251e1267. http://dx.doi.org/10.1111/j.1475-6773.2010.01135.x Low, G., & Molzahn, A. E. (2007). Predictors of quality of life in old age: a cross- validation study. Research in Nursing & Health, 30(2), 141e150. http://dx.doi. org/ 10.1002/nur.20178 Madigan, E. A., Gordon, N., Fortinsky, R. H., Koroukian, S. M., Pina, L., & Riggs, J. S. (2012). Predictors of functional capacity changes in a U.S. population of Medi- care home health care patients with heart failure. Archives of Gerontology & Geriatrics, 54(3), 300e306. http://dx.doi.org/10.1016/j.archger.2011.07.018 Montazeri, A. (2009). Quality of life data as prognostic indicators of survival in cancer patients: an overview of the literature from 1982 to 2008. Health and Quality of Life Outcomes, 7(102), 1e21. http://dx.doi.org/10.1186/1477-7525-7- 102 Murkofsky, R. L., & Alston, K. (2009). The past, present, and future of skilled home health agency care. Clinics in Geriatrics Medicine, 25(1), 1e17. http://dx.doi.org /10.1016/j.cger.2008.12.001 National Center for Health Statistics. (2005). Health, United States, 2005, with chartbook on trends in the health of Americans (pp. 322e323). Retrieved from. http://www.cdc.gov/nchs/data/hus/hus05 #094 Naylor, M. D., Brooten, D., Campbell, R., Maislin, G., McCauley, K. M., & Schwartz, S. (2004). Transitional care of older adults hospitalized with heart failure: a randomized, controlled trial. Journal of the American Geriatrics Society, 52(5), 675e684. http://dx.doi.org/10.1111/j.1532-5415.2004.52202.x Ory, M. G., & Cox, D. M. (1994). Forging ahead: linking health and behavior to improve quality of life in older people. Social Indicators Research, 33(1e3), 89e120. http://dx.doi.org/10.1007/BF01078959 Roberts, B. L. (1999). Activities of daily living: factors related to independence. In A. S. Hinshaw (Ed.), Handbook of clinical nursing research (pp. 563e578). New York: Sage. Scharpf, T. P., & Madigan, E. A. (2010). Functional status outcome measures in home health care patients with heart failure. Home Health Care Services Quarterly, 29(4), 155e170. http://dx.doi.org/10.1080/01621424.2010.534044 Shaughnessy, P. W., & Crisler, K. S. (2005). Effectiveness of a clinical feedback approach to improving patient outcomes: Handbook of home health care administration (4th ed.). Sudbury, MA: Jones and Bartlett Publishers. Shaughnessy, P. W., Hittle, D. F., Crisler, K. S., Powell, M. C., Richard, A. A., Kramer, A. M., et al. (2002). Improving patient outcomes of home health care: findings from two demonstration trials of outcome-based quality improvement. Journal of the American Geriatrics Society, 50(8), 1354e1364. http://dx.doi.org/10. 1046/j.1532-5415.2002.50356.x Shaughnessy, P. W., Schlenker, R. E., Crisler, K. S., Arnold, A. G., Powell, M. C., & Beaudry, J. M. (1996). Home care: moving forward with continuous quality improvement. Journal of Aging Society Policy, 7(3e4), 149e167. http://dx.doi.org/ 10.1300/J031v07n03_09 Skevington, S. M., Lotfy, M., & O’Connell, K. A. (2004). The world health organiza- tion’s WHOQOL-BREF quality of life assessment: psychometric properties and results of the international field trial. Quality of Life Research, 13(2), 299e310. http://dx.doi.org/10.1023/B. QURE.0000018486.91360.00 Tseng, S., & Wang, R. (2001). Quality of life and related factors among elderly nursing home residents in southern Taiwan. Public Health Nursing,18(5), 304e311. http:// dx.doi.org/10.1046/j.1525-1446.2001.00304.x Tullai-McGuinness, S., Madigan, E. A., & Fortinsky, R. H. (2009). Validity testing the outcomes and assessment information set (OASIS). Home Health Care Services Quarterly, 28(1), 45e57. Urden, L. D. (2001). Outcome evaluation: an essential component for CNS practice. Clinical Nurse Specialist, 15(6), 260e268. World Health Organization. (1998). Measuring quality of life: The world health or- ganization quality of life instruments The WHOQOL-100 and the WHOQOL-BREF. Retrieved from. http://www.who.int/mental_health/media/68 WHOQOL Group. (1994). Development of the WHOQOL: rationale and current status. International Journal of Mental Health, 23(3), 24e56. WHOQOL Group. (1995). The World Health Organization Quality of Life assessment (WHOQOL): Position paper from the World Health Organization. Special issue on health-related quality of life: What is it and how should we measure it? Social Science & Medicine, 41(10), 1403e1409. http://dx.doi.org/10.1016/j.archger.2007.11.005 http://dx.doi.org/10.1016/j.archger.2007.11.005 http://dx.doi.org/10.1007/s11136-011-9954-z http://dx.doi.org/10.1161/01.STR.31.8.2004 http://www.cms.hhs.gov/HomeHealthQualityInitis/Downloads/HHQIOverview/pdf http://www.cms.hhs.gov/HomeHealthQualityInitis/Downloads/HHQIOverview/pdf https://www.cms.gov/DataCompendium/14_2010_Data_Compendium.asp http://www.cms.gov/HomeHealthQualityInits/ http://www.lmpolicyresearch.com/sites/default/files/hhpps_literaturereview http://www.lmpolicyresearch.com/sites/default/files/hhpps_literaturereview http://dx.doi.org/10.1111/j.1547-5069.2008.00220.x http://dx.doi.org/10.1016/0021-9681(87)90014-2 http://dx.doi.org/10.1111/j.1945-1474.2004.tb00495.x http://dx.doi.org/10.2307/3563329 http://dx.doi.org/10.1111/j.1365-2648.2004.03247.x http://dx.doi.org/10.1080/13607861003801003 http://dx.doi.org/10.4040/jkan.2010.40.3.400 http://dx.doi.org/10.1007/s11136-011-0006-5 http://dx.doi.org/10.1007/s11136-011-0006-5 http://dx.doi.org/10.1111/j.1547-5069.2004.04017.x http://dx.doi.org/10.1111/j.1475-6773.2010.01135.x http://dx.doi.org/%2010.1002/nur.20178 http://dx.doi.org/%2010.1002/nur.20178 http://dx.doi.org/10.1016/j.archger.2011.07.018 http://dx.doi.org/10.1186/1477-7525-7-102 http://dx.doi.org/10.1186/1477-7525-7-102 http://dx.doi.org%20/10.1016/j.cger.2008.12.001 http://dx.doi.org%20/10.1016/j.cger.2008.12.001 http://www.cdc.gov/nchs/data/hus/hus05 #094 http://dx.doi.org/10.1111/j.1532-5415.2004.52202.x http://dx.doi.org/10.1007/BF01078959 http://dx.doi.org/10.1080/01621424.2010.534044 http://dx.doi.org/10.1046/j.1532-5415.2002.50356.x http://dx.doi.org/10.1046/j.1532-5415.2002.50356.x http://dx.doi.org/10.1300/J031v07n03_09 http://dx.doi.org/10.1300/J031v07n03_09 http://dx.doi.org/10.1023/B http://dx.doi.org/10.1046/j.1525-1446.2001.00304.x http://dx.doi.org/10.1046/j.1525-1446.2001.00304.x http://www.who.int/mental_health/media/68 Clinical Outcomes and Quality of Life of Home Health Care Patients Introduction Methods Study design Setting and samples Ethical considerations Measurements Clinical outcomes QOL General characteristics Data collection Data analysis Results General characteristics of participants Clinical outcomes between the start of care and discharge or 60 days QOL between the start of care and discharge or 60 days Determinants of change in QOL Discussion Clinical outcomes, ADL, and IADL QOL and functional outcomes Limitations Conclusion Conflict of interest Acknowledgments References RCT of NCM Diabetes Mgmt Original Research Diabetes Nurse Case Management in a Canadian Tertiary Care Setting: Results of a Randomized Controlled Trial Danni Li MD a, Tom Elliott MBBS a,b, Gerri Klein RN, CDE a,b, Ehud Ur MBBS a, Tricia S. Tang PhD a,* a Division of Endocrinology, Department of Medicine, University of British Columbia, Vancouver, British Columbia, Canada b BC Diabetes, Vancouver, British Columbia, Canada a r t i c l e i n f o Article history: Received 2 May 2016 Received in revised form 17 October 2016 Accepted 21 October 2016 Keywords: diabetes nurse case manager randomized controlled trial tertiary care setting diabetes distress psychosocial outcomes a b s t r a c t Objectives: To examine the effects of a 6-month nurse case manager (NCM) intervention compared to stan- dard care (SC) on glycemic control and diabetes distress in a Canadian tertiary-care setting. Methods: We recruited 140 adults with type 2 diabetes and glycated hemoglobin (A1C) levels >8%
(64 mmol/mol) from 2 tertiary care facilities and randomized them to: 1) a 6-month NCM intervention
in addition to SC or 2) SC by the primary endocrinologists. Assessments were conducted at baseline and
at 6 months. Primary outcomes included A1C levels and diabetes distress scores (DDS). Secondary out-
comes included body mass index, blood pressure, diabetes-related behaviour measures, depressive symp-
toms, self-motivation and perception of support.
Results: At the 6-month follow up, the NCM group experienced larger reductions in A1C levels of −0.73%
compared to the SC group (p=0.027; n=134). The NCM group also showed an additional reduction of −0.40
(26% reduction) in DDS compared to those in the SC group (p=0.001; n=134). The NCM group had lower
blood pressure, ate more fruit and vegetables, exercised more, checked their feet more frequently, were
more motivated, were less depressed and perceived more support. There were no changes and no group
differences in terms of body mass index, medication compliance or frequency of testing.
Conclusions: Compared to SC, NCM intervention was more effective in improving glycemic control and
reducing diabetes distress. It is, therefore, a viable adjunct to standard diabetes care in the tertiary care
setting, particularly for patients at high risk and with poor control.

© 2017 Canadian Diabetes Association.

Mots clés :
infirmière gestionnaire de cas de diabète
essai clinique à répartition aléatoire
cadre des soins tertiaires
détresse liée au diabète
critères de jugement psychosociaux

r é s u m é

Objectifs : Examiner les effets de l’intervention de 6 mois d’une infirmière gestionnaire de cas (IGC) par
rapport aux effets des soins courants (SC) sur la régulation de la glycémie et la détresse liée au diabète
dans le cadre canadien des soins tertiaires.
Méthodes : Nous avons recruté 140 adultes atteints du diabète de type 2 qui avaient des concentrations
d’hémoglobine glyquée (A1c)>8% (64 mmol/mol) de 2 établissements de soins tertiaires et les avons répartis
de manière aléatoire comme suit : 1) intervention de 6 mois d’une IGC en plus des SC ou 2) SC par les
endocrinologues traitants. Les évaluations ont été réalisées au début et après 6 mois. Les critères de jugement
principaux étaient les concentrations de l’A1c et les scores de la détresse liée au diabète (SDD). Les critères
secondaires étaient les suivants : l’indice de masse corporelle, la pression artérielle, les mesures du
comportement lié au diabète, les symptômes de dépression, la motivation personnelle et la perception
du soutien.
Résultats : Au suivi après 6 mois, le groupe IGC montrait des réductions plus grandes des concentrations
de l’A1c de −0,73% que celles du groupe SC (p=0,027; n=134). En plus, le groupe IGC montrait une réduction
des SDD de −0,40 (réduction de 26%) par rapport au groupe SC (p=0,001; n=134). Les adultes du groupe
IGC avaient une pression artérielle plus basse, mangeaient plus de fruits et de légumes, faisaient plus
d’exercice, vérifiaient leurs pieds plus fréquemment, étaient plus motivés, étaient moins dépressifs et
s’apercevaient d’un plus grand soutien. Il n’y avait aucun changement et aucune différence entre les groupes
en ce qui concerne l’indice de masse corporelle, l’observance thérapeutique ou la fréquence des analyses.

* Address for correspondence: Tricia S. Tang, PhD, University of British Columbia, 2775 Laurel Street, Suite 10211, Vancouver, British Columbia, Canada V5Z 1M9.
E-mail address: tricia.tang@vch.ca

Can J Diabetes 41 (2017) 297–304

Contents lists available at ScienceDirect

Canadian Journal of Diabetes
j o u r n a l h o m e p a g e :

w w w. c a n a d i a n j o u r n a l o f d i a b e t e s . c o m

1499-2671 © 2017 Canadian Diabetes Association.
The Canadian Diabetes Association is the registered owner of the name Diabetes Canada.
http://dx.doi.org/10.1016/j.jcjd.2016.10.012

Downloaded for Anonymous User (n/a) at University of Virginia from ClinicalKey.com by Elsevier on March 07, 2021.
For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved.

mailto:tricia.tang@vch.ca

http://dx.doi.org/10.1016/j.jcjd.2016.10.012

http://www.sciencedirect.com/science/journal/14992671

http://crossmark.crossref.org/dialog/?doi=10.1016/j.jcjd.2016.10.012&domain=pdf

Conclusions : Comparativement aux SC, l’intervention d’une IGC était plus efficace pour améliorer la régulation
de la glycémie et réduire la détresse liée au diabète. Par conséquent, il s’agit d’un complément viable aux
soins courants offerts aux diabétiques dans le cadre des soins tertiaires, particulièrement chez les patients
exposés à un risque élevé qui ont une régulation médiocre.

© 2017 Canadian Diabetes Association.

Introduction

The prevalence of diabetes is rising worldwide, primarily because
of an aging and increasingly obese population. In 2010, 2.7 million
(7.6%) people in Canada were diagnosed with diabetes, and this
number is projected to reach 4.2 million (10.8%) by 2020 (1). Dia-
betes and its associated complications are a significant burden on
the Canadian economy, costing $11.7 billion in 2010 and expected
to rise to $16 billion by 2020 (1).

Despite compelling evidence that tight glucose control can
prevent or delay complications (2), outcomes are poor, and improve-
ments are needed. For instance, among 3002 Canadian patients in
a primary care setting, Braga et al (3) found that 30%, 39% and 53%
achieved treatment targets for blood pressure (BP), glycated hemo-
globin (A1C) and cholesterol, respectively. Moreover, only 7%
achieved all 3 goals. Clearly, greater efforts are needed to help
patients improve diabetes-related health outcomes in Canada.

Among the various models investigated to improve diabetes care
delivery, case management has produced the most favourable evi-
dence (4,5). In fact, a meta-analysis of 11 different quality-
improvement strategies for diabetes care found that interventions
involving case-management strategies led to the greatest reduc-
tions in A1C levels (5). Case management encompasses the assess-
ment, implementation, coordination and monitoring of options and
services required to meet individual health needs (6). It can include
patients’ education, coaching, treatment adjustment, monitoring and
care coordination (7).

Several systematic and integrative reviews have shown diabe-
tes case management interventions to be effective in improving gly-
cemic control (6–10) by up to 0.89% (6). In addition, a study of 556
patients receiving care in a Veteran Affairs healthcare system found
that, compared to controls, a greater proportion of patients ran-
domized to a nurse case manager (NCM) intervention achieved the
collective treatment target for A1C levels, BP and low-density lipo-
protein levels (11).

Although there is overwhelming evidence supporting NCM
models in the treatment of diabetes, these studies have been con-
ducted predominantly in primary care and community-based set-
tings in the United States and Europe (6). In fact, of the 29 case
management studies in the meta-analysis of Welch et al (6), only
1 study was conducted with patients attending a tertiary care clinic
in Canada. That randomized controlled trial of 46 patients with dia-
betes found a significant reduction in A1C levels associated with a
telephone-based nursing intervention compared to standard care
conditions (12). However, the study recruited patients with both
type 1 and type 2 diabetes requiring insulin, and it focused on insulin
titration to the exclusion of other core self-management issues, such
as healthful eating, physical activity and psychosocial well-being.
To our knowledge, no studies of NCM-assisted patients with dia-
betes have been conducted in a Canadian tertiary care setting that
focus on comprehensive care of patients with type 2 diabetes only
who are being treated with oral agents and/or insulin.

The current study is the first randomized controlled trial (RCT)
to evaluate the impact of NCM intervention for patients with poorly
controlled type 2 diabetes who were recently discharged from 2 ter-
tiary care hospitals in Canada or referred by tertiary hospital-
affiliated endocrinologists. In addition, this Canadian-based study
is the first to include both a primary clinical outcome (A1C levels)
and a psychosocial outcome (diabetes distress).

Methods

Study design, setting and population

This study was approved by the University of British Columbia
and Providence Health Clinical Research Ethics Boards. It is an RCT
of a 6-month NCM intervention compared to standard care (SC). The
study was initiated in September 2012, enrollment was com-
pleted in July 2014, and follow up was completed in January 2015.
The protocol is viewable at https://clinicaltrials.gov/ct2/show/
NCT01659294.

The study was conducted at BCDiabetes.ca, based in the Gordon
and Leslie Diamond Health Care Centre. The centre is the main ter-
tiary care centre in Vancouver, British Columbia, and brings together
outpatient services at Vancouver General Hospital, including spe-
cialty clinics, along with medication education, physician teach-
ing clinics and research, at a single site.

Inclusion/exclusion criteria

To be eligible for the study, patients had to 1) have physician-
diagnosed type 2 diabetes; 2) be ≥18 years of age; (3) have A1C levels
≥8% and 4) to be able to read and write English. Patients were
excluded if they had previously worked with an NCM or had any
serious health conditions (e.g. terminal cancer), serious psychiat-
ric illness or self-reported excessive alcohol or illicit drug use that
would impede meaningful participation in the study.

Recruitment

Study participants were recruited using 2 streams: 1) patients
who had been recently discharged from 2 tertiary care hospitals in
Vancouver (Vancouver General Hospital or St. Paul’s Hospital) and
2) patients referred by endocrinologists affiliated with the 2 ter-
tiary care hospitals. If recruited from the hospital, the invitation to
participate in the study was made by a member of the treating endo-
crine team (endocrinology fellow or resident). The primary endo-
crinologist or the member of the team briefly described the study
and its eligibility criteria. The NCM contacted interested patients
and scheduled initial visits during which she described the study
in greater detail, obtained informed consent and conducted base-
line assessments. In the case of outpatient referrals, the invita-
tions came from the treating endocrinologist. These patients had
been referred to the treating endocrinologist by their community-
based family (primary) physician. All potential subjects who met
the study entry criteria were approached by treating endocrinolo-
gists to participate in the study. As such, participants were repre-
sentative of new diabetes referrals (with A1C levels >8.0) seen by
the referring endocrinologists.

Randomization process

Each participant was randomly assigned to the intervention or
control group by using a stratified permuted block randomization
scheme, with the endocrinologist being the sole stratification factor.
The permuted block aspect of the randomization scheme ensured
that treatment assignment remained balanced throughout the enrol-
ment period. Randomized assignments were completed in advance
and kept in individual, sealed, sequentially labelled envelopes that
were opened at the time of the randomization of each participant.

D. Li et al. / Can J Diabetes 41 (2017) 297–304298

Downloaded for Anonymous User (n/a) at University of Virginia from ClinicalKey.com by Elsevier on March 07, 2021.
For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved.

https://clinicaltrials.gov/ct2/show/NCT01659294

https://clinicaltrials.gov/ct2/show/NCT01659294

Nurse case manager intervention

Nurse case management was conducted by a single individual,
GK, a certified diabetes educator with a master’s degree in nursing
and 37 years of nursing experience. The nurse received informal
in-house diabetes training by working with 1 of the endocrinologists.

The general principle (although not the exact framework) used
by the NCM was the Empowerment and Self-Management model
developed by Funnell and Anderson (13).

NCM intervention was an adjunct to standard care. Partici-
pants in the NCM group received a 60-minute initial consultation
with the NCM, followed by contact via telephone or e-mail or in
person at a minimum of every other week initially and, thereaf-
ter, as each participant warranted. The purpose of follow-up contact
was to monitor participants’ diabetes control, risk factors and man-
agement goals and to recommend treatment additions or modifi-
cations where desirable. There were 3 core components of the NCM
intervention: 1) diabetes self-management education and support;
2) monitoring and algorithm-driven treatment adjustment; and 3)
care coordination with other health professionals.

Diabetes education
A brief education-needs assessment was made. All partici-

pants were asked to talk about their conditions and to say what they
felt had contributed to their diagnoses and what had helped them
to handle their diabetes and what had not. Participants were asked
to read information that was given and then to explain what they
had read in their own words.

During the initial consultation prior to randomization, the NCM
delivered diabetes education and provided supplementary written
literature to patients.

Broad topics included the process of diabetes, health behaviour
changes, healthful eating, physical activity and the role of medi-
cation, the meaning of A1C levels, the role of blood glucose moni-
toring, short- and long-term complications, stress and coping.
Whenever participants were prescribed sulfonylureas and/or insulin,
they were educated in the concept of purposeful blood glucose moni-
toring (target driven) and were given home blood glucose moni-
tors and 10 strips.

All of the NCM participants were asked what their target goals
were: blood glucose, exercise, weight (and smoking cessation if rel-
evant) and where they felt comfortable. Informal contracts between
participants and educators were discussed and implemented, either
in writing in the nursing notes or verbally, laying out a plan of what
the participants and nurses were to do.

Participants were given diabetes-related information in their
mother tongue where possible. Literacy was evaluated at consent,
when the participants were asked to read the consent form and
restate them in their own words. Some could not read English at
the level required for understanding and, therefore, were not given
written diabetes information. Picture-based materials from the Cana-
dian Diabetes Association were offered to those who appeared to
benefit from such materials. Those able to read at sufficient levels
of understanding were given written diabetes-related informa-
tion from the Canadian Diabetes Association and BCDiabetes.ca; free
literature from local big-chain pharmacies was offered; when appro-
priate, participants were given the book Understand Your Diabetes
(14).

Discussions where initiated about what was good information
and what was not so good. Other Internet sources were accessed
and discussed as ways of discovering what to believe and what not
to listen to.

Monitoring and algorithm-driven treatment adjustment
Participants randomized to the NCM cohort were reviewed by

the NCM at intervals determined and driven by the results of

purposeful blood glucose monitoring; glucose targets were deter-
mined by the referring endocrinologist. Participants were instructed
by the NCM to communicate their blood glucose monitoring results
by e-mail or phone. The NCM offered feedback and intervention
when they were considered appropriate. Blood pressure readings
and targets were discussed with participants; when medication
adjustments were considered desirable, the NCM communicated
directly with the endocrinologist. With respect to insulin adjust-
ment, the NCM independently adjusted insulin according to the fol-
lowing algorithms: 1) for once-daily basal insulin adjustment, target
fasting glucose levels were 5 mmol/L to 7 mmol/L. Patients were
encouraged to make daily adjustments in dose (default incre-
ments of 2 U for above-target and decrements of 4 U for below target,
no change for to-target values); 2) for twice-daily basal insulin,
fasting glucose targets were used to adjust evening insulin doses,
and evening blood glucose targets were used to adjust morning
insulin doses (same adjustment parameters as those for once-
daily basal insulin); 3) for prandial rapid insulin, participants were
asked to test before and 2 hours after meals; the before-meal targets
were 5 mmol/L to 7 mmol/L, and the 2-hour post meal targets were
5 mmol/L to 10 mmol/L. The NCM reviewed the concept of carbo-
hydrate counting and iterated the carbohydrate-to-insulin ratio.

Linkage to allied healthcare services
The NCM determined the need for adjunct interventions. When

indicated, and after consultation with endocrinologist and/or family
physicians, participants were referred to the following: outpa-
tient mental health clinics, smoking cessation clinics, physio-
therapy groups, nutrition counselling, social services and/or walking/
exercise/sports programs in the community.

Eye examination referrals were made to local expert optom-
etrists for all participants who had not had a retinal check in the
past year. Podiatric referrals were made when clinically indicated.
Specialist medical referrals for comorbid conditions were made as
necessary during the study by the treating endocrinologist.

With respect to communication between the participants and
the NCM, all participants were given the nurse’s contact informa-
tion (office phone, e-mail and, if on insulin, cell phone number).
The frequencies and dates and times of education and support were
agreed upon by the patients and the NCM. Patients who requested
more support accessed the NCM more often. The NCM would
follow up by phone or e-mail with patients new to insulin and set
up a contact schedule dependent upon patient preference and
reported blood sugar levels. Patients undergoing dose adjust-
ments contacted the NCM daily usually in the morning, with their
fasting blood sugar levels so as to gain assistance with dose adjust-
ments. However, once the patients were comfortable with the algo-
rithm, they managed their own dose adjustments independently
and contacted the NCM only for scheduled appointments. If the NCM
participant did not call at the expected time, the nurse called the
participant. After a third call by the nurse, all further calls were ini-
tiated by the participant unless for a further prearranged call.

With respect to diabetes-related distress and depression, dis-
cussion was initiated with all treatment participants about the asso-
ciation between depression and diabetes; answers to questions by
subjects in the study questionnaires were pursued. The nurse encour-
aged discussion about thoughts and feelings and about activities
the subjects enjoyed. Inquiries were made about alcohol and sub-
stance abuse and about access to counsellors or psychiatrists. Refer-
rals to psychiatrists were suggested and facilitated by the treating
endocrinologist, as necessary.

Standard care

Beyond the NCM’s arranging for the second and final visit at
6 months to collect study-specific data points, SC participants were

D. Li et al. / Can J Diabetes 41 (2017) 297–304 299

Downloaded for Anonymous User (n/a) at University of Virginia from ClinicalKey.com by Elsevier on March 07, 2021.
For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved.

managed entirely by the referring endocrinologist, who deter-
mined treatment parameters, made changes to treatment regi-
mens and scheduled follow-up appointments according to their usual
patterns of practice.

Participation and assessment

Participants in both the NCM group and the SC group com-
pleted baseline and 6-month assessments. A week prior to these
visits, participants were instructed to have their A1C levels mea-
sured at a self-selected laboratory. If no blood had been drawn prior
to the visit, blood was drawn on the day of the visit with the NCM,
and point-of-care A1C levels were measured at the time of the visit
so as to direct clinical decision making. At the NCM visit, partici-
pants had the following measured: BP, weight and height. Partici-
pants also completed a 20-minute self-report survey. Upon
completion, participants received a stipend of $20 for their time and
effort.

Outcomes and measurements

Primary outcomes included A1C levels and diabetes distress. The
A1C levels were measured at a participant-selected laboratory that
used standard high-performance liquid chromatography methods
on a machine calibrated to the national standard (mean 5.0%, top
of normal range = 6.0%). For participants who did not obtain blood
work at a laboratory prior to baseline and the 6-month assess-
ment, the NCM performed point-of-care testing using the Bayer
DCA2000+ Analyzer (Bayer, Keverkusen, Germany) (15). Diabetes
distress was assessed using the 17-item Diabetes Distress Scale (DDS)
developed by Polonsky and colleagues (16). The DDS measures emo-
tional distress and functioning as they relate to living with diabe-
tes. Responses are scored on a 6-point Likert scale (1 = no problem
to 6 = serious problem). A total score is derived by taking the mean
of all items. A score of <2 indicates low or no distress; a score of 2 through 2.9 indicates moderate distress; and a score of ≥3 indi- cates high distress. Secondary outcomes included clinical, behavioural and psycho- social outcome measures. BP was measured using an Omron Digital Blood Pressure Monitor HEM-907 (Kyoto, Japan). Two upper-arm readings were taken, and the average of the 2 was recorded. Height and weight were mea- sured using the Health o meter mechanical eye-level upright stan- dard physician’s scale with attached stadiometer (Sunbeam, Boca Raton, Florida, United States). Body mass indexes were calculated as weight in kilograms divided by height in metres squared. Self-management behaviours were assessed using items from the Summary of Diabetes Self-Care Activities Measure, revised (17). This instrument measures self-care behaviours, including diet, exer- cise, blood-sugar testing and foot care. Participants were asked to report the number of days in the past week (range, 0 to 7 days) during which they performed specific self-care practices. Greater numbers of days indicated better self-management. Medication adherence was assessed using the 4-item Morisky scale that assesses beliefs and behaviours associated with taking medications (18). Responses were scored on a dichotomous scale: 0 = no, and 1 = yes. A total score was calculated by adding up all items; lower scores indicated better adherence. Participant motivation was measured by the 13-item Patient Acti- vation Measure, which assesses participants’ self-reported knowl- edge, skills and confidence in managing their own health (19). Diabetes-specific social support was assessed by a 4-item scale devel- oped by Tang et al (20) that measures the amount of support and satisfaction with that support by family members, friends and the healthcare team. Depressive symptom severity was assessed using the 9-item PRIME-MD Patient Health Questionnaire (PHQ-9) (21,22). Demographic characteristics Demographic characteristics included age, gender, years diag- nosed with diabetes, ethnicity, marital status, education level, income and employment status. Sample size and power Based on the current literature, an A1C level difference of 0.5% or more was considered clinically significant. To achieve 80% power in a 1-sided 0.05-level t test, a sample size of 72 subjects per arm was considered necessary if the true difference in A1C levels between the NCM group and the SC group was 0.5%. This assumed a stan- dard deviation of the difference in A1C levels between baseline and 6 months of 1.2% and that it was common to both groups (23). Although DDS was established as an additional primary outcome at the onset of the study, it should be noted that sample size was calculated with only A1C levels in mind. Furthermore, many sec- ondary outcomes were included without changes in sample size. Recognizing the importance of well-powered studies, caution should thus be taken in interpreting the results. Statistical analysis Demographic characteristics of the sample are described as means and standard deviations if they are continuous and as counts and percentages if they are categorical. Baseline values for demo- graphic and clinical variables were compared between the 2 arms to identify potential confounding variables. Categorical variables were assessed by chi-square tests, with Monte Carlo-simulated p values when necessary. Because the primary outcomes (A1C levels and diabetes dis- tress) and the secondary outcomes are continuous measures, the assessment of whether the mean differences between the 2 groups were statistically significant was conducted using the Student t test, adjusting for baseline values. No adjustment for multiple testing was made. A modified intention-to-treat approach was used, in which data from all participants who provided baseline and follow-up out- comes were included and analyzed based on NCM-SC group assign- ments, regardless of whether each individual participant received the assigned treatment. Results Study sample There were 251 patients who met the inclusion criteria, and 140 patients who consented to participate in the study (Figure 1). Six patients did not complete baseline assessment. At the 6-month follow up, 130 patients completed A1C measurements, and 120 patients completed the survey (attrition rate 7% and 14%, respec- tively). Loss to follow up was not different in the 2 groups. Participants’ characteristics Table 1 presents the baseline demographic characteristics of the study’s population. No significant differences were found in the NCM and SC groups in terms of gender, ethnicity, martial status, educa- tion level, employment status or household income (all p values >0.08). Although various ethnic groups were involved in the study,
about half were Caucasian. The mean age of the participants was
57 years, and the mean number of years of living with diabetes was
11; 54% were married, 58% were male, 31% had high school edu-
cations or less, 54% were employed and 48% reported an annual
income of less than $50 000.

D. Li et al. / Can J Diabetes 41 (2017) 297–304300

Downloaded for Anonymous User (n/a) at University of Virginia from ClinicalKey.com by Elsevier on March 07, 2021.
For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved.

Primary outcomes

Glycemic control
At baseline, glycemic control was inadequate for both the NCM

and the SC groups (Figure 2, Table 2). At the 6-month follow up,
both groups had achieved significant reductions in mean A1C levels

(from 10.45%±2.13% to 7.72%±1.43; p<0.01, and from 10.52%±2.08 to 8.46%±2.29; p<0.01, respectively). The NCM group experienced significantly larger reductions in A1C levels (−0.73%) (Figure 2, A) compared to the SC group (p=0.027; n=134). Diabetes distress At baseline, both the NCM and the SC groups had diabetes dis- tress in the low to moderate category. At the 6-month follow up, the NCM group had significantly lower mean DDS scores com- pared to those found at the start of the intervention (from 1.90±0.77 to 1.37±0.38; p<0.01). This implies that the DDS score was reduced from near moderate to low. The SC group showed no change in DDS scores (from 1.92±0.84 to 1.78±0.91; p=0.23). Participants in the NCM group showed an additional reduction of −0.40 (26%) in DDS scores (Figure 2B) compared to those in the SC group (p=0.001; n=134), and all had low levels of diabetes distress at the end of the study. Secondary outcomes Clinical outcomes In the NCM group, the average systolic and diastolic BP declined by 8 mm Hg (p=0.02) and 4 mm Hg (p=0.04), respectively, from base- line to 6 months (Table 2). The SC group did not show any improve- ment in BP levels. There was no change in body mass indexes in either group; they were in the obese category at baseline and at the 6-month follow up. Figure 1. Study flowchart. A1C, HbA1c; GP, general practitioner; NCM, nurse case man- agement; SC, standard care. Table 1 Demographic information concerning the modified intent-to-treat population All N=134 SC n=64 NCM n=70 Age Mean (SD) 57.43 (11.01) 58.23 (10.87) 56.69 (11.16) Years with diabetes Mean 10.84 11.84 9.91 (SD) (8.24) (8.10) (8.33) Gender Male 78 (58%) 37 (58%) 41 (59%) Female 56 (42%) 27 (42%) 29 (41%) Ethnicity Arabic 3 (2%) 2 (3%) 1 (1%) Black 4 (3%) 2 (3%) 2 (3%) Chinese 19 (14%) 5 (8%) 14 (20%) First Nations 5 (4%) 4 (6%) 1 (1%) South Asian 15 (11%) 9 (14%) 6 (9%) White 65 (49%) 29 (45%) 36 (51%) Other 23 (17%) 13 (20%) 10 (14%) Marital status Never married 27 (20%) 11 (17%) 16 (23%) Married 72 (54%) 33 (52%) 39 (57%) Living with partner 3 (2%) 3 (5%) 0 (0%) Separated/divorced 20 (15%) 12 (19%) 8 (12%) Widowed 11 (8%) 5 (8%) 6 (9%) n missing 1 (1%) 0 1 (1%) Education Less than high school 12 (9%) 9 (14%) 3 (4%) High school 29 (22%) 11 (17%) 18 (26%) College/technical 32 (24%) 15 (24%) 17 (24%) University 41 (31%) 22 (35%) 19 (27%) Graduate degrees 19 (14%) 6 (10%) 13 (19%) n missing 1 (1%) 1 (1%) 0 Income $20,000 30 (22%) 20 (31%) 10 (14%) $20,000–$29,000 14 (10%) 6 (9%) 8 (11%) $30,000–$39,000 17 (13%) 7 (11%) 10 (14%) $40,000–$49,000 16 (12%) 10 (16%) 6 (9%) $50,000–$59,000 9 (7%) 5 (8%) 4 (6%) $60,000–$69,000 11 (8%) 5 (8%) 6 (9%) $70,000 37 (28%) 11 (17%) 26 (37%) Employment Full time 60 (45%) 26 (41%) 34 (49%) Part time 12 (9%) 6 (9%) 6 (9%) Unemployed 9 (7%) 5 (8%) 4 (6%) Homemaker 4 (3%) 2 (3%) 2 (3%) In school 1 (1%) 0 (0%) 1 (1%) Retired 35 (26%) 17 (27%) 18 (26%) Other 13 (10%) 8 (12%) 5 (7%) NCM, nurse case management; SC, standard care. Note: No significant differences between randomized groups; all p values >0.05.

Figure 2. Changes in A1C levels and DDS scores in 6 months in the NCM and SC groups.
Sample means with 95% CIs for each treatment group at baseline and after 6 months
are shown. A1C, HbA1c; DDS, diabetes distress score; NCM, nurse case manage-
ment; SC, standard care.

D. Li et al. / Can J Diabetes 41 (2017) 297–304 301

Downloaded for Anonymous User (n/a) at University of Virginia from ClinicalKey.com by Elsevier on March 07, 2021.
For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved.

Behavioural outcomes
NCM participants consumed more fruit and vegetables (an

increase of 1 day of eating more than 4 servings of fruit and veg-
etables in a week; p<0.01); exercised more (an increase of 1 day of more than 30 minutes of exercise in a week; p<0.01) and checked their feet more often (an additional increase of 1 day per week for feet checks; p=0.049) compared to the SC group at the 6-month follow up. Both groups were similar in number of blood sugar checks in a week (4 to 5 checks per week) and medication compliance (score of 1 on the 4-item Morisky scale, which indicated good medica- tion compliance), at the baseline and at the 6-month follow up. Psychosocial outcomes At baseline, both groups had mild depression, based on the PHQ-9 scores. At the 6-month follow up, NCM participants had fewer depressive symptoms (an additional reduction of 3 points; p<0.01), with average score indicating minimal depression. There was no change in the SC group. NCM participants also reported a greater amount of support received and were more motivated. Although both groups, at baseline, had mean Patient Activation Measure scores that fell into the stage 3 category (stage 3 indicates taking actions for diabetes self-care in normal circumstances), NCM participants were able to move to stage 4 (p<0.01), which indicates ability for self-care in circumstances of physical illness or mental distress. Other observations During the course of the study, 9 patients in the NCM group and 11 patients in SC group were hospitalized. There were no hospi- talizations for hypoglycemia or acute hyperglycemia in either group. Table 2 Primary and secondary outcomes: changes from baseline All N=134 SC n=64 NCM n=70 Difference in changes between groups, adjusted for baseline (95% CI) p value Primary outcomes Change in A1C −2.41 −2.06 −2.72 −0.731 0.027 (2.57) (2.58) (2.54) (−1.377, −0.085) q=4 q=3 q=1 Change in DDS −0.35 −0.15 −0.54 −0.403 0.001 (0.83) (0.92) (0.68) (−0.629, −0.176) q=14 q=6 q=8 Secondary outcomes Clinical outcome measures BMI 0.05 −0.20 0.28 0.564 0.190 (2.53) (3.15) (1.78) (−0.283, 1.411) q=17 q=8 q=9 SBP −3.58 −0.12 −6.71 −8.059 0.021 (21.24) (20.03) (21.96) (−14.861, −1.257) q=16 q=8 q=8 DBP −3.50 −0.64 −6.08 −4.911 0.014 (13.00) (13.92) (11.62) (−8.803, −1.020) q=16 q=8 q=8 Behavioural outcome measures Fruit and vegetable intake 0.39 −0.26 1.02 1.112 0.010 (2.95) (2.92) (2.86) (0.274, 1.950) q=15 q=6 q=9 Fat intake 0.08 0.07 0.08 −0.761 0.056 (2.45) (2.94) (1.9) (−1.541, 0.019) q=15 q=6 q=9 Physical activity 0.4 −0.17 0.94 1.277 0.007 (3.23) (2.83) (3.5) (0.359, 2.195) q=14 q=6 q=8 Specific exercise 0.26 −0.29 0.77 1.319 0.005 (3.12) (2.97) (3.19) (0.418, 2.220) q=14 q=6 q=8 Foot exam 1.34 0.62 2.02 1.006 0.049 (3.56) (3.44) (3.56) (0.004, 2.008) q=14 q=6 q=8 Blood glucose monitoring 0.88 0.90 0.85 −0.259 0.528 (2.94) (3.12) (2.79) (−1.069, 0.551) q=14 q=6 q=8 Medication adherence 0.01 0.05 −0.03 −0.130 0.441 (1.17) (1.33) (1.01) (−0.464, 0.203) q=15 Q=7 Q=8 Psychosocial outcome measures Patient activation measure 3.16 1.16 5.03 6.364 <0.001 (10.86) (11.37) (10.09) (3.335, 9.393) q=14 q=6 q=8 PHQ-9 −0.41 1.09 −1.82 −2.954 <0.001 (4.67) (4.42) (4.48) (−4.381, −1.527) q=14 q=6 q=8 Amount of support 0.32 −0.05 0.68 0.504 0.004 (1.29) (1.30) (1.18) (0.169, 0.838) q=14 q=6 q=8 Satisfaction with support 0.21 0.00 0.4 0.297 0.075 (1.18) (1.25) (1.09) (−0.030, 0.625) q=15 q=7 q=8 A1C, glycated hemoglobin; BMI, body mass index; DBP, diastolic blood pressure; DDS, Diabetes Distress Scale; PHQ-9, 9-item Patient Health Questionnaire; SBP, systolic blood pressure. Note: Data are mean (SD); q = number of missing observations. Differences between groups are adjusted for baseline, and the associated p values are calculated using linear regression. D. Li et al. / Can J Diabetes 41 (2017) 297–304302 Downloaded for Anonymous User (n/a) at University of Virginia from ClinicalKey.com by Elsevier on March 07, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. Discussion This is the first RCT of a NCM intervention involving patients with type 2 diabetes in a Canadian tertiary care setting that demon- strated that NCM was effective in improving glycemic control and reducing diabetes distress. Furthermore, this intervention also led to improvements in other clinical, behavioural and psychosocial outcomes. Given that our sample was drawn from a tertiary care setting, participants were more likely to have poorer glycemic control and were at higher risk than the general population of persons with type 2 diabetes. Therefore, it is not surprising this group would be more receptive and amenable to an NCM intervention (12,24). This was a behavioural-intervention study, so it is difficult to delineate which component of the intervention worked best. It is interesting that medication compliance and glucose testing did not change over the course of the study for either the intervention or the control group. These findings suggest that improvements in life- style behaviours, such as dietary patterns and physical activity, may have been the mechanism of change. Alternatively, frequent medi- cation adjustment may also account for A1C level and blood pres- sure reductions in the intervention group; however, that variable was not measured in this study. The rate of diabetes distress is high in patients with type 2 dia- betes. Previous research has found NCM interventions to improve psychosocial outcomes, such as depression (25), but only 1 has reported a reduction in diabetes distress (26). This the first study of NCM presence conducted in a Canada-based specialty care setting to demonstrate reduction in diabetes distress and also improve- ments in depressive symptoms, patient motivation and perceived social support. Diabetes distress is associated with worse glycemic control (27). In fact, Pandit et al (28) found that, compared to minimally dis- tressed patients, highly distressed patients had higher levels of A1C, diastolic BP and LDL cholesterol. Diabetes self-management edu- cation has been shown to decrease diabetes distress, such that a reduction of 10 points on the Problem Areas in Diabetes (PAID) scale was associated with a clinically significant reduction in A1C levels of 0.55% (29). It is possible that improvements in distress and depres- sion also contributed to A1C-level reductions. Future studies using structural equation modelling should be conducted to examine this question. Limitations to this study need to be acknowledged. The study budget relied upon blood laboratory analysis being reimbursed by the BC Ministry of Health; all blood draws were subject to province- specific guidelines. These guidelines support 3 monthly A1C tests, but a lipid panel is recommended only annually. Consequently, not all physicians ordered lipid panels at both baseline and 6 months and, indeed, some were resistant to this level of testing. Given this inconsistency, we could not include lipid profiles or any other clini- cal outcomes that are measured only annually. Finally, we did not build a formal cost analysis into the study; therefore, we are not able to provide any cost-effectiveness data. However, such an NCM model has already been adopted nationwide in a primary care setting in Germany and has proven to improve survival time, lower mean daily hospital and mean daily drug costs (30). Conclusions Nurse case management was an effective model for improving A1C levels and reducing diabetes distress in a tertiary care setting. Therefore, it is a viable adjunct to standard care, particularly for the inadequately controlled patients with type 2 diabetes who are referred for specialist care. Acknowledgements The authors acknowledge the contributions made by the patients, the physicians, the nurse case manager and the office staff, in addi- tion to the staff at the collaborating hospitals, who have made this work possible and specially acknowledge Harlan Campbell and Darby Thompson for their help with statistical analysis. Author Disclosures This study was supported by the British Columbia Endocrine Research Foundation (itself supported by a grant from Sanofi- Aventis Canada). Clinical trial registry #NCT01659294, clinicaltrials.gov. Dr. Tom Elliott has received funding for giving lec- tures and symposia and attending advisory boards for the follow- ing companies: Sanofi, NovoNordisk, Boehringer-Ingelheim and Astra-Zeneca. Dr. Ehud Ur has received funding for giving lectures and symposia and attending advisory boards for the following com- panies: Sanofi, NovoNordisk, Boehringer-Ingelheim, Merck and Astra-Zeneca. Author Contributions TT, TE and EU were the principal investigators; DL, TT TE and EU contributed to study’s conception and design, data analysis and interpretation; GK was the study coordinator and collected data for analysis; DL and TE prepared the manuscript for submission, and TT, GK and EU reviewed and edited the manuscript prior to its submission. References 1. Canadian Diabetes Association. Diabetes: Canada at the tipping point. http:// www.diabetes.ca/publications-newsletters/advocacy-reports/diabetes-canada -at-the-tipping-point. Accessed October 1, 2013. 2. The Diabetes Control and Complications Trial Research Group. The effect of inten- sive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 1993;329:977–86. 3. Braga M, Casanova A, Teoh H, et al. Treatment gaps in the management of car- diovascular risk factors in patients with type 2 diabetes in Canada. Can J Cardiol 2010;26:297–302. 4. Saxena S, Misra T, Car J, et al. Systematic review of primary healthcare inter- ventions to improve diabetes outcomes in minority ethnic groups. J Ambul Care Manage 2007;30:218–30. 5. Shojania KG, Ranji SR, McDonald KM, et al. Effects of quality improvement strat- egies for type 2 diabetes on glycemic control: A meta-regression analysis. JAMA 2006;296:427–40. 6. Welch G, Garb J, Zagarins S, et al. Nurse diabetes case management interven- tions and blood glucose control: Results of a meta-analysis. Diabetes Res Clin Pract 2010;88:1–6. 7. Pimouguet C, Le Goff M, Thiebaut R, et al. Effectiveness of disease-management programs for improving diabetes care: A meta-analysis. CMAJ 2011;183:E115– 27. 8. Sutherland D, Hayter M. Structured review: Evaluating the effectiveness of nurse case managers in improving health outcomes in three major chronic diseases. J Clin Nurs 2009;18:2978–92. 9. Norris SL, Nichols PJ, Caspersen CJ, et al. The effectiveness of disease and case management for people with diabetes: A systematic review. Am J Prev Med 2002;22:15–38. 10. Joo JY, Huber DL. An integrative review of case management for diabetes. Prof Case Manag 2012;17:72–85. 11. Ishani A, Greer N, Taylor BC, et al. Effect of nurse case management compared with usual care on controlling cardiovascular risk factors in patients with dia- betes: A randomized controlled trial. Diabetes Care 2011;34:1689–94. 12. Thompson DM, Kozak SE, Sheps S. Insulin adjustment by a diabetes nurse edu- cator improves glucose control in insulin-requiring diabetic patients: A ran- domized trial. CMAJ 1999;161:959–62. 13. Funnell M, Anderson R. Empowerment and self-management of diabetes. Clin Diabetes 2004;223:123–7. 14. Bernard S, Cormier C, Josse RG, et al. Understand your diabetes and live a healthy life. Montreal: Rogers Publishing, 2013. D. Li et al. / Can J Diabetes 41 (2017) 297–304 303 Downloaded for Anonymous User (n/a) at University of Virginia from ClinicalKey.com by Elsevier on March 07, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. http://clinicaltrials.gov http://www.diabetes.ca/publications-newsletters/advocacy-reports/diabetes-canada-at-the-tipping-point http://www.diabetes.ca/publications-newsletters/advocacy-reports/diabetes-canada-at-the-tipping-point http://www.diabetes.ca/publications-newsletters/advocacy-reports/diabetes-canada-at-the-tipping-point http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0015 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0015 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0015 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0015 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0020 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0020 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0020 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0025 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0025 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0025 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0030 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0030 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0030 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0035 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0035 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0035 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0040 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0040 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0040 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0045 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0045 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0045 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0050 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0050 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0050 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0055 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0055 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0060 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0060 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0060 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0065 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0065 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0065 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0070 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0070 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0075 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0075 15. Arsie MP, Marchioro L, Lapolla A, et al. Evaluation of diagnostic reliability of DCA 2000 for rapid and simple monitoring of HbA1c. Acta Diabetol 2000;37: 1–7. 16. Polonsky WH, Fisher L, Earles J, et al. Assessing psychosocial distress in diabe- tes: Development of the diabetes distress scale. Diabetes Care 2005;28: 626–31. 17. Toobert DJ, Hampson SE, Glasgow RE. The summary of diabetes self-care activi- ties measure: Results from 7 studies and a revised scale. Diabetes Care 2000;23:943–50. 18. Morisky DE, Green LW, Levine DM. Concurrent and predictive validity of a self-reported measure of medication adherence. Med Care 1986;24: 67–74. 19. Hibbard JH, Stockard J, Mahoney ER, Tusler M. Development of the Patient Acti- vation Measure PAM: Conceptualizing and measuring activation in patients and consumers. Health Serv Res 2004;39:1005–26. 20. Tang TS, Brown MB, Funnell MM, Anderson RM. Social support, quality of life, and self-care behaviors among African Americans with type 2 diabetes. Dia- betes Educ 2008;34:266–76. 21. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: Validity of a brief depression severity measure. J Gen Intern Med 2001;16:606–13. 22. Spitzer RL, Kroenke K, Williams JB. Validation and utility of a self-report version of PRIME-MD: The PHQ primary care study. Primary Care Evaluation of Mental Disorders. Patient Health Questionnaire. JAMA 1999;282:1737– 44. 23. Monteiro AP. Psychometric properties of a Russian version of the Social Support Questionnaire SSQ6 in Eastern European immigrants. Contemp Nurse 2011;39:157–62. 24. Fanning EL, Selwyn BJ, Larme AC, DeFronzo RA. Improving efficacy of diabetes management using treatment algorithms in a mainly Hispanic population. Dia- betes Care 2004;27:1638–46. 25. Couch C, Sheffield P, Gerthoffer T, et al. Clinical outcomes in patients with type 2 diabetes managed by a diabetes resource nurse in a primary care practice. Proc (Bayl Univ Med Cent) 2003;16:336–40. 26. Gabbay RA, Anel-Tiangco RM, Dellasega C, et al. Diabetes nurse case manage- ment and motivational interviewing for change DYNAMIC: Results of a 2-year randomized controlled pragmatic trial. J Diabetes 2013;5:349–57. 27. Fisher L, Mullan JT, Arean P, et al. Diabetes distress but not clinical depression or depressive symptoms is associated with glycemic control in both cross- sectional and longitudinal analyses. Diabetes Care 2010;33:23–8. 28. Pandit AU, Bailey SC, Curtis LM, et al. Disease-related distress, self-care and clini- cal outcomes among low-income patients with diabetes. J Epidemiol Commu- nity Health 2014;68:557–64. 29. Leyva B, Zagarins SE, Allen NA, Welch G. The relative impact of diabetes dis- tress vs depression on glycemic control in Hispanic patients following a dia- betes self-management education intervention. Ethn Dis 2011;21:322–7. 30. Drabik A, Buscher G, Sawicki PT, et al. Life-prolonging of disease management programs in patients with type 2 diabetes is cost-effective. Diabetes Res Clin Pract 2012;95:194–200. D. Li et al. / Can J Diabetes 41 (2017) 297–304304 Downloaded for Anonymous User (n/a) at University of Virginia from ClinicalKey.com by Elsevier on March 07, 2021. For personal use only. No other uses without permission. Copyright ©2021. Elsevier Inc. All rights reserved. http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0080 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0080 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0080 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0085 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0085 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0085 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0090 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0090 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0090 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0095 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0095 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0095 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0100 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0100 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0100 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0105 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0105 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0105 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0110 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0110 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0115 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0115 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0115 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0115 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0120 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0120 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0120 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0125 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0125 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0125 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0130 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0130 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0130 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0135 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0135 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0135 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0140 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0140 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0140 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0145 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0145 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0145 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0150 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0150 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0150 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0155 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0155 http://refhub.elsevier.com/S1499-2671(16)30114-9/sr0155 Diabetes Nurse Case Management in a Canadian Tertiary Care Setting: Results of a Randomized Controlled Trial Introduction Methods Study design, setting and population Inclusion/exclusion criteria Recruitment Randomization process Nurse case manager intervention Diabetes education Monitoring and algorithm-driven treatment adjustment Linkage to allied healthcare services Standard care Participation and assessment Outcomes and measurements Demographic characteristics Sample size and power Statistical analysis Results Study sample Participants' characteristics Primary outcomes Glycemic control Diabetes distress Secondary outcomes Clinical outcomes Behavioural outcomes Psychosocial outcomes Other observations Discussion Conclusions Acknowledgements Author Disclosures Author Contributions References