1.efficacy of an empowerment program for taiwanese patients

Applied Nursing Research 28 (2015) 366–373

Contents lists available at ScienceDirect

Applied Nursing Research

j ourna l homepage: www.e lsev ie r .com/ locate /apnr

Original Article

Efficacy of an empowerment program for Taiwanese patients with type 2
diabetes: A randomized controlled trial
Mei-Fang Chen, RN, PhD a,1, Ruey-Hsia Wang, RN, PhD b,⁎, Kuan-Chia Lin, PhD c,2,Hsiu-Yueh Hsu, RN, PhD d,3, Shu-Wen Chen, RN, PhD e,4

a Department of Nursing, National Tainan Junior College of Nursing, Tainan City 700, Taiwan, R.O.C.b College of Nursing, Kaohsiung Medical University, Kaohsiung 807, Taiwan, R.O.C.c Department of Healthcare Administration and Management, National Taipei University of Nursing and Health Sciences, Peitou District, Taipei City, Taiwan, R.O.C.d Hsin Sheng Junior college of medical care and management, Taoyuan, Taiwan, R.O.C.e College of Nursing, Central Taiwan University of Science and Technology, Taichung City 40601, Taiwan R.O.C.

a b s t r a c ta r t i c l e i n f o

Funding: No specific funding was received for this stuConflict of interest: The authors declare no conflict ofContribution: Study design: MFC, RHW; data collect

manuscript preparation: MFC, RHW, HYH, SWC.⁎ Corresponding author at: College of Nursing, Kaohsiun

Chuan 1st Road, Kaohsiung 807, Taiwan, R.O.C. Tel.: +886E-mail addresses: [email protected] (M.-F. Che

(R.-H. Wang), [email protected] (K.-C. Lin), [email protected] (S.-W. Chen).

1 Tel.: +886 62110900x280.2 Tel.: +886 228227101x6122.3 Tel.: +886 34117578x110.4 Tel.: +886 422391647.

http://dx.doi.org/10.1016/j.apnr.2014.12.0060897-1897/© 2015 Elsevier Inc. All rights reserved.

Please cite this article as: Chen, M-F., et al.,controlled trial, Applied Nursing Research (20

Article history:
Received 19 March 2014Revised 24 December 2014Accepted 25 December 2014Available online xxxx
Keywords:EmpowermentDiabetesRandomized control trialGlycosylated hemoglobinSelf-care behaviors

Aim: To examine the efficacy of an empowerment programon glycosylated hemoglobin, self-care behaviors, self-efficacy, and quality of life in Taiwanese patients with type 2 diabetes.Background: Although empowerment interventions have proven beneficial in western populations, they arerarely applied in Taiwanese populations.Methods: A randomized controlled trial was conducted. The experimental group (n = 33) participated in a3-month empowerment program for motivating patient self-awareness, assessing the causes of problems,goal setting, individual self-care plan development, and checking whether the goal is reached (MAGIC).The control group (n = 32) received routine clinical care.Results: The experimental group had significantly decreased glycosylated hemoglobin at 3 months after theend of the intervention and significantly improved self-care behaviors, self-efficacy, and quality of life at theend of the intervention and at 3 months after the end of the intervention.
Conclusions: An empowerment program effectively improves diabetes control in Taiwanese patients withtype 2 diabetes.
© 2015 Elsevier Inc. All rights reserved.

1. Introduction

The high prevalence, highmortality and hugemedical costs of type 2diabetes are now urgent health problems in many countries. Annually,diabetes affects more than 347million people worldwide and causes al-most 3 million deaths (World Health Organization, 2013). In Taiwan, arecent study showed that 9.2% of the population has been diagnosedwith diabetes and that diabetes has become the fifth leading cause ofdeath (TaiwanNational Department of Health, 2013). Diabetes accountsfor 11.5% of overall health care spending. Of the total amount spent fortreating diabetes, three-fourths is spent on treating diabetes-related

dy.interest.ion and analysis: MFC, KCL and

gMedical University, 100, Shih-73121101x2641.n), [email protected]@meiho.edu.tw (H.-Y. Hsu),

Efficacy of an empowerment15), http://dx.doi.org/10.1016

complications (Taiwanese Association of Diabetes Educators, 2006). Toreduce the impact of diabetes in Taiwan, developing appropriate diabe-tes care programs is essential.

Traditional diabetes education focuses on patient compliance withthe recommendations of healthcare professionals. The "compliance"strategy makes patients feel powerless towards their diabetes control(Anderson & Funnell, 2010). The empowerment strategy applies a par-ticipatory process that enables individuals to achieve a sense of controlover their lives (Herbert, Gagnon, Rennick, & O'loughlin, 2009) and re-duces feelings of powerlessness (Falk-Rafael, 2001). According to theWorld Health Organization (2006), empowerment is an appropriatestrategy for people with diabetes.

Empowerment can be considered a process or an outcome. As a pro-cess, empowerment enables people to choose to take control over andmake decisions about their lives (Chen, Wang, Chin, Chen, & Chen,2011; Falk-Rafael, 2001). Diabetes empowerment refers to actionsprovided by healthcare professionals to assist people with diabetes intaking control of their lives (Anderson & Funnell, 2010). Diabetesempowerment processes can include awareness, action and reflectionphases (Chen et al., 2011a; Chen, Wang, & Tang, 2011). Awareness rais-ing is the initial phase of empowering patients (Falk-Rafael, 2001).Healthcare professionals should help patients recognize their rightsand capabilities to make decisions that affect their own health (Chen

program for Taiwanese patients with type 2 diabetes: A randomized/j.apnr.2014.12.006

http://crossmark.crossref.org/dialog/?doi=10.1016/j.apnr.2014.12.006&domain=pdf

http://dx.doi.org/10.1016/j.apnr.2014.12.006

mailto:[email protected]






http://www.sciencedirect.com/science/journal/08971897


367M-F. Chen et al. / Applied Nursing Research 28 (2015) 366–373

et al., 2011a; Chen et al., 2011b; Falk-Rafael, 2001). Patients who recog-nize that they can make decisions are likely to take responsibility fortheir diabetes care. In the action phase, mutual participation, open com-munication, and providing necessary information are emphasized(Chen et al., 2011a; Chen et al., 2011b). Mutual participation empha-sizes that healthcare professionals act as ‘facilitators’whoencourage pa-tients actively to participate in their diabetes control, such as addressingquestions, setting goals, and making decisions (Anderson & Funnell,2010; Chen et al., 2011a; Chen et al., 2011b). Open communication is es-sential for building equal partnerships between healthcare profes-sionals and patients. In an equal partnership, patients feel free to sharetheir experiences and opinions with their healthcare professionals(Anderson & Funnell, 2010). Healthcare professionals should also pro-vide the information that patients need to integrate new knowledgeand skills in coping with diabetes (Chen et al., 2011a; Chen et al.,2011b). Mutually evaluating the goal between healthcare professionalsand patients can help patients modify actions and goals; therefore, re-flection is necessary at the end of the empowerment process (Falk-Rafael, 2001).

An empowered person is expected to demonstrate positive behav-ioral, physical and psychosocial outcomes (Anderson & Funnell, 2010).A widely used indicator of the outcome of diabetes control is HbA1clevels since lowHbA1c level reduces the risk of diabetes-relatedmortal-ity and complications (UK Prospective Diabetes Study Group, 1998).HbA1c is a key physical indicator of diabetes control (Mulcahy et al.,2003). Quality of life (QOL), an intangible, subjective perception oftotal life experiences (Plummer & Molzahn, 2009), is considered a psy-chosocial outcome of diabetes control (Mulcahy et al., 2003). Patientswith type 2 diabetes are expected to perform good self-care behaviorsbecause self-care behaviors are significantly associated with HbA1clevel (Yu et al., 2013). Self-efficacy is defined as the perceived self-confidence to plan and take actions (Bandura, 1997). Traditionally,self-efficacy is considered a psychosocial outcome of empowerment(Yu et al., 2013). Self-efficacy is positively associated with self-care be-haviors and QOL in patients with type 2 diabetes (Liu, Tai, Hung,Hsieh, & Wang, 2010). Therefore, when empowering patients with dia-betes, HbA1c, self-care behaviors, self-efficacy, and QOL can be consid-ered important outcomes.

Previous studies have shown that empowerment programs effec-tively improve HbA1c (Pibernik-Okanovic, Prasek, Poljicanin-Filipovic,Pavlic-Renar, & Metelko, 2004), self-care behaviors (Kuo, Tsay, & Yang,2002), self-efficacy (Kuo et al., 2002), and QOL (Lowe, Linjawi, Mensch,

Empowerment

Theoretical

Framework

Diabetes

Empowerment

Program

Framework

Empowerment Process

Empowerment Intervention

Awareness phase

Raising

awareness

Action phase

Mutual participation

Open communication

Providing necessary

information

Motivating

patient

self-awareness

Assessing the causes of

problems

Goal setting

Individual self-care plan

development

Fig. 1. Empowerment theoretical framework and

Please cite this article as: Chen, M-F., et al., Efficacy of an empowermentcontrolled trial, Applied Nursing Research (2015), http://dx.doi.org/10.1016

James, & Attia, 2008) in people with type 2 diabetes. Nevertheless, mostempowerment interventions tend to focus on assessing the effective-ness on specific outcomes. An effective diabetes care program shouldcomprehensively target behavioral, physical and psychosocial outcomes(Mulcahy et al., 2003).

An empowered patient needs a healthcare provider who is receptiveto the patient sharing in decision making (Anderson & Funnell, 2010).However, in the partriarchal society of Taiwan, patients tend to obeythe instructions of healthcare professionals and are generally lessempowered than patients in other countries. Therefore, whenTaiwanese patients have difficulty implementing the lifestyle changesor dietary changes recommended by healthcare professionals, theytend to blame themselves rather than collaboratively seeking a solutionin their daily life. Patientsmay feel powerless and begin to play a passiverole in their diabetes care. Powerlessness has been reported in patientswithdiabetes in Taiwan (Liu et al., 2010). Furthermore, only 34.5%of pa-tients with type 2 diabetes had HbA1c less than 7% in Taiwan (Yu et al.,2013). An effective empowerment program would encourageTaiwanese patientswith type 2 diabetes to be proactive in their diabetescontrol. Since the effect of empowermentmay differ across populations(Yip, 2004), examining the effectiveness of an empowerment programspecifically in Taiwanese patients with type 2 diabetes is crucial. Thepurpose of this study was to evaluate the efficacy of the empowermentprogram on HbA1c, self-care behaviors, self-efficacy, and QOL in pa-tients with type 2 diabetes in Taiwan. The primary outcome is the out-come of greatest importance. Data on secondary outcomes are used toevaluate additional effects of the intervention (Outcome, 2014). SinceHbA1c is the most important indicator that may be directly associ-ated with complications, HbA1c was considered the primary out-come. Self-care behaviors, self-efficacy and QOL were consideredsecondary outcomes in this study. Fig. 1 shows the framework ofthe empowerment intervention.

2. Methods

2.1. Design

This randomized controlled trial was performed from May, 2010 toFebruary, 2011. Participants were randomly assigned to either experi-mental or control groups. A systematic review indicated that most em-powerment interventions last at least 6 weeks (Chen et al., 2011b).Therefore, the experimental group received a 3-month empowerment

Empowerment Outcomes

Positive behavioral, physical,

and psychosocial outcomes

Empowerment Outcomes of

Diabetes Control

Primary outcome

Physical indicator: HbA1c

Secondary outcome

Behavioral indicator:

Self-care behaviors

Psychosocial indicator:

Self-efficacy

Quality of life

Reflection phase

Evaluating the

goal and

further

modification

Checking

whether the

goal is

reached

diabetes empowerment program framework.



368 M-F. Chen et al. / Applied Nursing Research 28 (2015) 366–373

program whereas the control group received routine clinical care. Toevaluate the immediate and short-term efficacy of the empowermentprogram, outcomesweremeasured at baseline (T0), at the endof the in-tervention (T1), and at 3 months after the end of the intervention (T2).The HbA1C reflected the glucose control in the past 3 months. We hy-pothesized that outcomes would not significantly differ between T0and T1. The HbA1c was measured at T0 and T2. Demographic anddisease-related characteristics were only collected at T0. This studywas approved by the Human Experiment and Ethics Committee ofKaohsiung Medical University (KMUH-IRB-990120).

2.2. Recruitment

Patients with type 2 diabetes were recruited by convenience sam-pling from the endocrinology clinics of two hospitals in Kaohsiung,Taiwan that provided similar diabetes care services. The inclusioncriteria were (1) diagnosis with type 2 diabetes, (2) clear mental statusand ability to communicate, (3) age between 40 and 70 years, (4) liter-ate and able to write in Chinese, (5) independent and able to perform

Signed consent form

Baseline measures:

1. Questionnaire: Personnel C

Self-care Scale; Chinese DE

2. Blood HbA1C

Experimental group (n=36)

Empowerment program

Lost to follow-up (n=3:

transferred to another hospital,

n=2; traveled abroad, n=1)

Measurement at 1 week post- intervention:

Diabetes Self-care Scale, Chinese DES


Measurement at 3 months post-intervention:

Diabetes Self-care Scale, Chinese DES


Randomized (n

Allocatio

Follow-U

Analysi

Enrollment Approached participa(n=76)

Fig. 2. Flowchart of procedure for recruiting study participants, intervention and measu


self-care, and (6) HbA1c level higher than 8% during the previous3 months. Patients who had any history of cancer, other chronic condi-tions (e.g., congestive heart failure), or mental disease were excluded.Physicians screened potential participants for eligibility and explainedthe purposes and procedure of the study to them. Eligible patientswho agreed to participate were then referred to the researchers. The re-searchers informed the participants that the study was voluntary, thattheir decision to participate would not affect their treatment, and thatthey could withdraw from the study at any time. Each participantgave written informed consent before the intervention.

The G*Power (Germany; version 3.1.1) software was used to esti-mate the required sample size (Fual, Erdfelder, Lang, & Buchner,2007). A pilot study indicated that the effect sizes for HbA1c at3 months after the end of the intervention between experimental(n = 10) and control groups (n = 10) were 0.72 (Chen, 2011). Givenan effect size of 0.72 identified in the pilot study, an α level of 0.05,and a power of 0.8, the study required 32 participants in each group.Assuming a 10% attrition rate, 36 participants were required foreach group.

(n=72)Refused (n=4)

haracteristics; Diabetes

S; Chinese DQOL

Control group (n=36)

Routine clinical care

Lost to follow-up (n=4: traveled

abroad, n=2; transferred

to another hospital, n=1;

unknown status, n=1)

, Chinese DQOL, Blood HbA1c


, Chinese DQOL, Blood HbA1c


=72)

n

p

s

nts

rement. DES, Diabetes Empowerment Scale; DQOL, Diabetes Quality of Life Scale.




Out of 76 subjects invited to participate, 72 agreed to participate.After the baseline measurements, a nurse who did not participate ineither data collection or data analysis randomized participants into ex-perimental or control group by block randomization with a block sizeof six. Overall, 33 participants in the experimental group (11 and 22 ineach hospital) and 32 participants in the control group (11 and 21 ineach hospital) completed the study at T2. Retention rates for the exper-imental group and control group were 91.7% and 88.9% at T2, respec-tively. Fig. 2 presents the flowchart of the procedure for recruitingparticipants, intervention, and measurements.

2.3. Intervention

After a comprehensive literature review, a 5-step MAGIC empower-ment program was developed. The five steps were: Motivating patientself-awareness, Assessing the causes of problems, Goal setting, Individ-ual self-care plan development, and Checking whether the goal wasreached. Table 1 summarizes the details of the program.

Table 1Five-step empowerment program.

Steps and contents

Step 1: Motivating patient self-awareness· Discuss with participants their HbA1C, blood pressure, and lipids levels and

the comparison with normal range.· Discuss personal values regarding risk of short- and long-term complications.· Encourage participants to share their experiences in performing self-care

behaviors, including diet, exercise, and medication behaviors.· Facilitate participants to reflect how their self-care behaviors affect their

HbA1C, blood pressure, lipids, and the possible complications.· Facilitate participants in increasing their awareness of self-care behavior

problems and in setting priorities for solving problems.

Step 2:Assessing the causes of problems· Encourage participants to explore their barriers and support their practice of

self-care behaviors. Sample questions:“What obstacles prevent you from practicing self-care behaviors?”“Which of these obstacles can be overcome?”“Under what conditions are you willing to practice your self-care behaviors?”“What support do you have when practicing self care behaviors?”

Step 3: Goal setting· Review ADA-recommended goals regarding HbA1C, blood pressure and

blood lipids.· Discuss the meanings of short-, medium- and long-term goals.· Encourage participants to share their expectations about diabetes care.· Encourage participants to establish achievable goals with the assistance of

the intervener.

Step 4: Individual self-care plan development· Encourage participants to choose themost suitable plan for achieving their goals.· Encourage participants to incorporate self-care plans in daily life activities.· Mutually establish a self-care behavior plan by applying Specific, Measurable,

Achievable, Realistic, Time (SMART) strategies.· Access necessary diabetes-related health information, resources and services

provided by healthcare providers.· If needed, invite significant family members to discuss their feelings about

diabetes and how they can help participants regulate their self-care behaviors.· Demonstrate how participants can use the self-monitoring checklist to record

their self-care behaviors.

Step 5: Checking whether the goal is reached· Guide participants to reflect on their experiences in self-care behaviors and on

the goals reached. Sample questions:“Please share your experience in performing self-care behaviors during thepast week.”“What barriers did you overcome?”“What goals did you reach?”“What did you learn from this experience?”· Praise participants who achieve their goals and apply steps 1 to 4 to resolve

the next problem. For participants who did not achieve their goals, discuss thereasons why goals were not reached, and explore solutions by applying steps 2to 4.


According to the regulations of the TaiwanBureau of National HealthInsurance, all patients with diabetes are scheduled to make monthlyclinical visits. For the convenience of participants in the experimentalgroup, a face-to-face interviewwas performed in a private room duringeach clinical visit. Three weekly telephone interviews were performedin the 3 weeks after the monthly visit. In the face-to-face interview,the intervener facilitated participants in sharing their self-care experi-ences. Because diet, exercise, and medications are strongly correlatedwith HbA1c (Mulcahy et al., 2003), self-care behaviors associated withdiet, exercise, and medications were addressed in the empowermentprogram. Participants were asked to identify themost troubling aspectsof living with diabetes (step 1). The intervener then assisted partici-pants in assessing the causes of identified problems (step 2), settinggoals (step 3), and developing individual self-care plans (step 4). Duringthe following 3weeks, the intervener performedweekly telephone inter-views to follow up on the performance of the participants and to providefurther instructions as needed. At the next face-to-face interview whenpatients visited the clinic, the intervener applied step 5 to checkwhetherthe goal set in the previous face-to-face interview had been reached. Ifthe goal had not been reached, the intervener repeated steps 2 to 4. Ifthe goal had been reached, the intervener applied steps 1 to 4 to addressanother problem expressed by the participant. Then, three telephone in-terviewswereperformed to followup again. The procedurewas repeatedacross the 3-month intervention. The duration of the face-to-face inter-views was 60–70 min, and that of the telephone interviews was30–40min. By the end of the intervention, each participant had complet-ed three face-to-face interviews and nine telephone interviews. All inter-views were performed by a single researcher who was a certifieddiabetes educator and who had completed a 2-day empowerment edu-cation course offered by the TaiwaneseAssociation ofDiabetes Educators.

To ensure the fidelity of participants, all participants in the experi-mental group received a booklet including eight sections: the contractwith the intervener; record of medical examination; record of self-care behaviors practice; self-evaluation of the impacts of diabetes andself-care behaviors; the priorities for problem of self-care behavior;self-care behaviors plans; health-related diabetes information and use-ful resources; and a self-monitoring checklist for diet, exercise andmed-ications. Participants in the experimental groupwere asked to completethe records in the booklet along with the progress of the intervention.All participants in the experimental group completed 12 interviewsand completed the booklet recording.

The participants in the control group received routine clinical careduring their monthly clinical visits. In routine clinical care, participantsreceived diabetes education from diabetes educators according to theguidelines of exercise and diet modifications developed by the TaiwanBureau of National Health Insurance. Each face-to-face interview lastedabout 30min. Participants also received an educational pamphlet aboutdiabetes care.

2.4. Measurements

Blood samples and structured questionnaires were used tocollect data.

2.4.1. Blood samplingAfter the participants completed the questionnaires at T0 and T2, a

research assistant drew blood samples by venipuncture. HbA1cwas an-alyzed by laboratory examiners who were blinded to the two groups ofparticipants. The HbA1c was measured by high performance liquidchromatography in a hospital laboratory certifiedby the TaiwanAccred-itation Foundation (reference range, 4.0%–6.0%). The research assistantcollected the HbA1c value from medical records.

2.4.2. Self-care behaviorsSelf-care behaviors were assessed using three subscales of the

Diabetes Self-care Scale (Hurley & Shea, 1992), which assessed diet (6



Table 2Summary of personal characteristics by group.

Variables Experimental group Control group χ2 / t p

n (%) / Mean (SD) n (%) / Mean (SD)

SexMale 16 (44.4) 17 (47.2) 0.06 .813Female 20 (55.6) 19 (52.8)

Age (years) 62.12 (7.51) 61.72 (8.79) 0.21 .643Marital statusMarried 33 (91.7) 34 (94.4) 0.32 .573Single or widowed 3 (8.3) 2 (5.6)

Education levelNo formal education 5 (13.9) 8 (22.2) 1.52 .467Elementary or juniorhigh school

23 (63.9) 18 (50.0)

Senior high school or above 8 (22.2) 10 (27.8)Employment statusEmployed 12 (33.3) 15 (41.7) 0.53 .465Unemployed 24 (66.7) 21 (58.3)

Tobacco useYes 5 (13.9) 3 (8.3) 0.56 .453No 31 (86.1) 33 (91.7)

Alcohol useYes 2 (5.6) 4 (11.1) 0.73 .394No 34 (94.4) 32 (88.9)

Duration of DM (years) 8.12(3.25) 7.45(3.83) 0.76 .451DM complicationsYes 27 (75.0) 24 (66.7) 0.61 .437No 9 (25.0) 12 (33.3)

Treatment methodOHA 18 (50.0) 22 (61.1) 1.53 .466Insulin 2 (5.6) 3 (8.3)OHA + Insulin 16 (44.4) 11 (30.6)

DM, diabetes mellitus; OHA, oral hypoglycemic agent.


items), exercise (4 items), and medication (3 items) behaviors. Re-sponses were rated from 1 point (complete non-performance) to 5points (complete performance). Higher scores indicated better self-care behaviors. The total possible score ranged from 13 to 65. In a previ-ous study of a Taiwan population with diabetes, the content validityindex, test–retest reliability, and Cronbach's alpha obtained for the Dia-betes Self-care Scale were 0.92, 0.83, and 0.86, respectively (Liu et al.,2010). The Cronbach's alpha was 0.87 in this study.

2.4.3. Self-efficacyAn 8-item Chinese version of the Diabetes Empowerment Scale

(DES, Lin, 2005)was used to assess self-efficacy. TheDESwas developedto measure the self confidence of patients in managing diabetes. Eachitem was rated from 1 point (strongly disagree) to 5 points (stronglyagree). Higher scores indicated better self-efficacy. The total possibletotal score ranged from 8 to 40. The Chinese version of the DES obtaineda Cronbach's alpha of 0.86 in a Taiwan population with diabetes (Lin,2005). In this study, Cronbach's alpha was 0.91.

2.4.4. Quality of lifeQuality of life was measured with a Chinese version of the Diabetes

Quality of Life (DQOL) Scale developed by Liu et al. (2010). The DQOLsubscales include satisfaction (15 items), impact (20 items), anddiabetes-related worry (7 items). In the satisfaction subscale, eachitem was rated from 1 point (very dissatisfied) to 5 points (very satis-fied). In the impact and diabetes-related worries subscale, each itemwas rated from 1 point (always) to 5 points (never). Higher scores indi-cated better QOL. The total possible score ranged from 42 to 210. In theTaiwan population analyzed in Liu et al. (2010), the Cronbach's alphaand test–retest reliability of the Chinese version of the DQOL Scalewere 0.91 and 0.83 respectively. In this study, the Cronbach's alpha ofthe DQOL Scale was 0.90.

2.4.5. Personal characteristicsPersonal characteristics included demographics and disease-related

characteristics. Demographics included sex, age, marital status, educa-tion level, employment status, tobacco use, and alcohol use. Disease-related characteristics included duration of diabetes, history of compli-cations, and treatment methods.

2.5. Data collection procedure

At T0, a trained research assistant in the endocrinology clinic visitedby the participant administered the questionnaire and drew blood sam-ples in a private room. For each participant, completing the question-naire and drawing the blood sample required approximately 25 min.Depending on the group assignment, each participant then received ei-ther the empowerment programor routine clinical care. Thedata collec-tion procedures at T1 and T2 were the same as T0 except that no bloodsamples were drawn at T1, i.e., at T1, the questionnaire was adminis-tered, but no blood samples were drawn; at T2, the questionnaire wasadministered, and blood samples were drawn. The research assistantwas blinded to the group assignments of the participants and did notprovide medical services to any participants throughout the durationof the study.

2.6. Analysis

Data were analyzed using SPSS Version 17 (SPSS, Inc., Chicago,IL). Student t-test and Chi-square test were used to examine groupdifferences in personal characteristics and outcome variables.Paired-sample t-tests were used to compare within-group differ-ences in outcome variables between T0 and T1, between T0 and T2and between T1 and T2. The generalized estimating equations(GEE) model was used to identify the independent effect of the em-powerment program, to consider within-person variability, and to


account for correlated data resulting from repeated measurementsacross different time points and multiple observations of the sameindividual (Zeger & Liang, 1986). The GEEmodel is considered an ap-propriate method for repeated measurement analysis. In compari-son with maximum likelihood approaches, the GEE model isconsidered robust against the working correlation structure, whichis assumed to be correct for within-subject correlations. Theinteracting effects of group and time on outcome variables were ex-amined by GEE model. To minimize type 1 error, a p value b .001 wasconsidered statistically significant.

3. Results

3.1. Comparison of personal characteristics and outcome variables at T0between the experimental and control groups

Before the intervention, the experimental and control groups did notsignificantly differ in terms of demographics or disease-related charac-teristics (Table 2). This indicated homogeneity between the twogroups; therefore, personal characteristics were not adjusted in theGEE models. Furthermore, the experimental and control groups didnot significantly differ in HbA1c levels, self-care behaviors, self-efficacy, or QOL at T0 (Table 3).

3.2. Difference in outcome variables between experimental and controlgroup at T1 and T2

Table 3 shows that, at T1 and T2, self-care behaviors, self-efficacyand QOL were significantly higher in the experimental group than inthe control group. However, HbA1c levels did not significantly differat T2.



Table 3Distribution of HbA1c, self-care behaviors, self-efficacy, and quality of life and comparisons between and within groups at T0, T1, and T2.

Variables T0 T1 T2 T1–T0 T2–T0 T2–T1

Mean (SD) Mean (SD) Mean (SD) t-value (p) t-value (p) t-value (p)

HbA1CExperimental group 9.24 (1.00) – 8.37 (1.03) – −5.90 (b .001) –

Control group 8.95 (0.69) – 8.73 (0.80) – −1.82 (.078) –

t-value (p) 1.34 (.186) −1.53 (.130)Self-care behaviors

Experimental group 40.70 (6.65) 53.55(6.92) 53.97 (5.81) 10.45(b .001) 12.54 (b .001) 0.65 (.518)Control group 42.03 (8.35) 45.28(8.72) 43.84 (8.28) 3.35 (.002) 2.15 (.040) −1.96 (.059)t-value (p) −0.71 (.478) 4.24 (b .001) 5.69 (b .001)

Self-efficacyExperimental group 23.15 (3.84) 30.03 (3.50) 30.76 (3.67) 10.76 (b .001) 10.25 (b .001) 1.98 (.056)Control group 24.84 (4.83) 25.84 (4.93) 26.06 (4.49) 1.80 (.082) 1.70 (.098) 0.38 (.710)t-value (p) −1.57 (.122) 3.96 (b .001) 4.62 (b .001)

Quality of lifeExperimental group 145.61 (17.62) 172.85 (11.18) 173.94 (11.38) 8.58 (b .001) 9.39 (b .001) 1.20 (.241)Control group 152.88 (14.61) 159.88 (13.39) 159.31 (13.10) 4.53 (b .001) 3.35 (.002) −0.37 (.713)t-value (p) −1.81 (.075) 4.25 (b .001) 4.81 (b .001)

T0, baseline; T1, the end of the intervention; T2, 3 months after the end of the intervention.–, Not applicable because HbA1C was not measured at T1.

Table 4GEE analysis of changes in HbA1c, self-care behaviors, self-efficacy, and quality of life be-tween experimental and control group.

Variables B SE p value

HbA1CIntercept 8.95 0.17 b .001Group (EG vs. CG) 0.29 0.23 .222T2 vs. T0 −0.23 0.14 .115Time*Group overall b .001EG*(T2 vs. T0) vs CG*(T2 vs. T0) −0.64 0.20 b .001

Self-care behaviorsIntercept 42.03 1.33 b .001Group (EG vs. CG) −1.33 1.86 .474Time overall b .001T1 vs. T0 3.25 0.86 b .001T2 vs. T0 1.81 1.15 .116Time*Group overall b .001EG*(T1 vs. T0) vs CG*(T1 vs. T0) 9.60 1.21 b .001EG*(T2 vs. T0) vs CG*(T2 vs. T0) 11.46 1.62 b .001

Self- efficacyIntercept 24.84 0.75 b .001Group (EG vs. CG) −1.69 1.05 .108Time overall b .001T1 vs. T0 1.00 0.53 .061T2 vs. T0 1.22 0.71 .084Time*Group overall b .001EG*(T1 vs. T0) vs CG*(T1 vs. T0) 5.88 0.75 b .001EG*(T2 vs. T0) vs CG*(T2 vs. T0) 6.39 0.99 b .001

Quality of lifeIntercept 152.88 2.43 b .001Group (EG vs. CG) -7.27 3.40 .033Time overall b .001T1 vs. T0 7.00 2.11 .001T2 vs. T0 6.44 2.69 .017Time*Group overall b .001EG*(T1 vs. T0) vs CG*(T1 vs. T0) 20.24 2.97 b .001EG*(T2 vs. T0) vs CG*(T2 vs. T0) 21.90 3.77 b .001

GEE, generalized estimating equation; T0, baseline; T1, the end of the intervention; T2, 3 -

months after the end of the intervention; EG, experimental group; CG, control group.


3.3. Difference of outcome variables between T0 and T1, T0 and T2, T1 andT2 within experimental and control groups

As shown in Table 3, participants of the experimental group signifi-cantly improved their HbA1c levels from T0 and T2. Participants of theexperimental group significantly improved their self-care behaviors,self-efficacy, and QOL from T0 to T1 as well as from T0 to T2.

In terms of comparisons between T1 and T2, no significant differ-ences were found in self-care behaviors, self-efficacy, and QOL. In thecontrol group, only QOL significantly improved from T0 to T1. However,HbA1c did not significantly differ between T0 and T2. Self-care behav-iors and self-efficacy did not significantly differ between T0 and T1 aswell as between T0 and T2.

3.4. Differencing of changing amount of outcome variables betweenexperimental and control groups at T1 and T2

The group and time interaction and adjusted corresponding out-come variables at T0 were then included in the GEE analysis. Table 4shows that, in terms of time main effect, self-care behaviors, and QOLsignificantly differed over the three time points. In terms of groupmain effect, no group effects were found in any outcome variables.However, significant interacting effects of time and groupwere revealedin HbA1c, self-care behaviors, self-efficacy, and QOL. Increasingamounts of self-care behaviors (BT1–T0 = 9.60, p b .001; BT2–T0 =11.46, p b .001), self-efficacy (BT1–T0 = 5.88, p b .001; BT2–T0 = 6.39,p b .001), and QOL (BT1–T0 = 20.24, p b .001; BT2–T0 = 21.90, p b .001)from T0 to T1 and from T0 to T2 in the experimental group werestatistically significantly larger than those in the control group.Decreasing amounts of HbA1c in the experimental group were statisti-cally significantly larger than those in the control group from T0 to T2(BT2–T0 = −0.64, p b .001).

4. Discussion

Compared to the control group, the experimental group significantlyimproved in HbA1c, self-care behaviors, self-efficacy, andQOL at T2. Theempowerment program can effectively simultaneously improve physi-cal, behavioral, and psychosocial outcomes. The results of this studysupported the efficacy of an empowerment program in a Chinese popu-lation with type 2 diabetes.

HbA1c significantly decreased by 0.87% from T0 to T2. We suggestthat the empowerment program had positive effect on decreasingHbA1c. An earlier study of 3- to 6-month empowerment interventions


performed in western populations reported HbA1c decreases rangingfrom 0.39% to 0.71% (Pibernik-Okanovic et al., 2004). Empowerment in-terventions performed in different countries have shown similar effica-cy in decreasing HbA1c. The UK Prospective Diabetes Study Group(1998) reported that a 1%decrease in HbA1c is associatedwith a 21% re-duction in diabetes-related death and a 37% reduction in microvascularcomplications. The empowerment program in the current study obtain-ed a 0.87% reduction in HbA1c, which may be sufficient to reduce




diabetes-related death and complications. HbA1c can be affected bymedication and body weight. The medications of the two groups werenot changed during intervention (data not shown). However, furtherstudies can follow up the change of body weight of participants and as-sess the influence on HbA1c.

The self-care behaviors in the experimental group significantly im-proved from T0 to T1 and from T0 to T2, which indicated the empower-ment program had positive effect on improving self-care behaviors inpatients with type 2 diabetes. In the experimental group, self-care be-haviors improved at T1 and the improvements were maintained at T2.A previous study found the self-care behaviors of patients with diabeteswere improved at 8 weeks after the end of the empowerment interven-tion (Kuo et al., 2002). Combined with the result of this study, the em-powerment program in this study had early effect on self-carebehaviors. Furthermore, the improvement in self-care behaviors in theexperimental group was sustained for up to 3 months after the end ofthe intervention. The empowerment program encouraged participantsto make autonomous, informed decisions about their self-care behav-iors. By playing an active role in their self-care practices, the participantsin the experimental group developed sustained self-care behaviors. Inthe control group, self-care behaviors increased at T1; however, theydecreased at T2. Participants in the control group received diabetes ed-ucation once amonth andwere providedwith educational pamphlets ofdiabetes control. A study also found thatmonthly visits to clinics and re-ceiving educational pamphlets improved the self-care behaviors of par-ticipants at the beginning, but the effect was not sustained at 3 monthsafter the end of the intervention (Kuo et al., 2002). In this study, self-care behaviors were measured by three selected subscales of DiabetesSelf-Care Scale. The procedure might change the instrument properties.To further validate the efficacy of empowerment program on self-carebehaviors, additional studies need to apply a comprehensive scale tomeasure self-care behaviors.

In comparison with T0, self-efficacy was significantly improved atT1 and T2 in the experimental group but not in the control group. AtT1 and T2, the increases in self-efficacy in the experimental groupwere significantly larger than those in the control group. Similarfindings have been reported previously (Kuo et al., 2002). Since mas-tery of an experience increases self-efficacy (Bandura, 1997), the im-proved self-efficacy in the experimental group likely resulted fromtheir repeated success in problem solving processes, which then in-creased their confidence. In contrast, participants in the controlgroup did not practice problem-solving processes, which may resultin no effect on their self-efficacy.

TheQOLwas significantly improved at T1 and T2 in both experimen-tal and control groups. However, the increased amount in the experi-mental group was significantly larger than that of the control group atT1 and T2. The results of this study are consistent with those of a previ-ous study (Lowe et al., 2008). The empowerment intervention of thisstudy encouraged participants to explore methods of coping with theburden of diabetes care. Therefore, improvement of QOL at T1 and T2in the experimental group was significantly larger than that of the con-trol group. Self-care behaviors directly influence the QOL (Mulcahyet al., 2003). The improvement of self-care behaviors may contributeto the improvement of QOL.

Compared to the possible total score, the scores for self-care behav-iors, self-efficacy andQOLwere atmedium–high levels at T1,which lim-ited the potential improvement from T1 to T2. Since self-care behaviorsand self-efficacy both can affect HbA1c, the increases in self-care behav-iors and self-efficacy at T1 may have contributed to the decreases inHbA1c at T2. Effect sizes below 0.32; from 0.33 to 0.55; and above0.55 are defined as small, medium, and large effect sizes, respectively(Cohen, 1988). The effect sizes of self-care behaviors, self-efficacy, andQOL were 1.42, 1.15, and 1.19, respectively at T2. The empowermentprogram had a large effect size in self-care behaviors, self-efficacyand QOL, which was consistent with an earlier systematic review(Chen et al., 2011a; Chen et al., 2011b). These data suggest that an


empowerment program is useful for improving behavioral and psycho-social outcomes.

Each participant in the experimental group completed 12 interviewsand made all required entries in the booklet. The high intensity inter-vention and fidelity might contribute to the positive outcome of thisempowerment program. However, it might limit the application ofempowerment program on clinical settings. Further studies of theempowerment program are needed to determine how decreasing thefrequency or duration of face-to-face interviews affects the outcomevariables. Minimizing the frequency and duration of the interviewswould simplify delivery of the program in primary care or diabetes cli-nic settings. The retained sample sizes of experimental and controlgroups were 33 and 32 respectively, which satisfied the sample estima-tion at the beginning of the study. The post-hoc power estimate forHbA1c at T2 was 0.99, which suggests that the retained sample sizewas adequate for examining the efficacy of the empowerment programfor lowering HbA1c.

Some limitations of this study should be considered. All participantswere recruited from two hospitals in southern Taiwan, which mightlimit the generalizability of the study. The absence of an attention con-trol group indicated lack of blinding the participants of the controlgroup. The measurements of secondary outcomes were based on self-reported scales, whichmight influence the estimation of efficacy of em-powerment program. Further studies with a more rigorous design areneeded to evaluate the effects of the empowerment program in differ-ent samples. An individual-based empowerment program was deliv-ered in this study. According to social learning theory (Bandura,1997), observation of others is effective in improving self-efficacy andbehaviors. We can examine the efficacy of group-based empower-ment intervention in the future. Additionally, the efficacy of the in-tervention was evaluated only 3 months after the end of theintervention. To assess the long-term effects of the empowermentprogram, further studies with longer follow-up periods such as6 months or 1 year are also needed. Finally, tests of the efficacy ofthe program in future studies can include body mass index as an in-dicator of diabetes control.

5. Conclusion and clinical implications

The empowerment program effectively reduced HbA1c levels andimproved self-care behaviors, self-efficacy, and QOL in patients withtype 2 diabetes in two hospitals in Taiwan. Empowerment interventioncould be of benefit for patients with type 2 diabetes across different cul-tures. The study could provide a reference for healthcare providers andresearcherswhen designing empowerment programs for patients livingwith type 2 diabetes.

Conflict of interest

Noauthors of this studyhave personal, professional, orfinancial con-flicts of interest to declare.

Funding

None.

Ethical approval

Ethical approval was given by the Kaohsiung Medical Universityin Taiwan.

References

Anderson, R. M., & Funnell, M. M. (2010). Patient empowerment: Myths and mis-conceptions. Patient Education and Counseling, 79(3), 277–282, http://dx.doi.org/10.1016/j.pec.2009.07.025.

Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W.H. Freeman.


http://dx.doi.org/10.1016/j.pec.2009.07.025

http://refhub.elsevier.com/S0897-1897(15)00041-5/rf0010



Chen, M. F. (2011). The effects of an empowerment-based education program for patients withtype 2 diabetes. Published doctoral dissertation Taiwan: Kaohsiung Medical University.

Chen, M. F., Wang, R. H., Chin, C. C., Chen, C. Y., & Chen, C. Y. (2011a). The content ofempowering with diabetes patients: A qualitative research. The New Taipei Journal,13(2), 9–20.

Chen, M. F., Wang, R. H., & Tang, S. M. (2011b). The application and effectiveness of anempowerment intervention among diabetes patients: A systematic review. TaiwanJournal of Public Health, 30(2), 1–17.

Cohen, J. (1988). Statistical power analysis for the behavioural sciences. Hillsdale, NJ:Lawrence Erlbaum Associates, 19–74.

Falk-Rafael, A. R. (2001). Empowerment as a process of evolving consciousness: A modelof empowerment caring. Advances Nursing Science, 24(1), 1–16, http://dx.doi.org/10.1097/00012272-200109000-00004.

Fual, F., Erdfelder, E., Lang, A. G., & Buchner, A. (2007). G*Power 3: A flexible statisticalpower analysis program for social, behavioral, and biomedical sciences. BehaviorResearch Methods, 39(2), 175–191.

Herbert, R. J., Gagnon, A. J., Rennick, J. E., & O'loughlin, J. L. (2009). A systematic re-view of questionnaires measuring health-related empowerment. Research andTheory for Nursing Practice, 23(2), 107–132, http://dx.doi.org/10.1891/1541-6577.23.2.107.

Hurley, A. C., & Shea, C. A. (1992). Self-efficacy: Strategy for enhancing diabetes self-care.Diabetes Educator, 18(2), 146–150, http://dx.doi.org/10.1177/014572179201800208.

Kuo, P. L., Tsay, S. L., & Yang, C. (2002). The effectiveness of empowerment training on theHbA1c, self-efficacy and self-care among diabetes patients. Formosan Journal ofMedicine, 6(6), 848–857.

Lin, C. C. (2005).Development and testing of the diabetes self-management instrument. Pub-lished doctoral dissertation Michigan: State University of Michigan.

Liu, M. Y., Tai, Y. K., Hung, W. W., Hsieh, M. C., & Wang, R. H. (2010). Relationships be-tween emotional distress, empowerment perception and self-care behavior andquality of life in patients with type 2 diabetes. The Journal of Nursing, 57(2), 49–60.

Lowe, J., Linjawi, S., Mensch, M., James, K., & Attia, J. (2008). Flexible eating and flexibleinsulin dosing in patients with diabetes: Results of an intensive self-managementcourse. Diabetes Research and Clinical Practice, 80(3), 439–443, http://dx.doi.org/10.1016/j.diabres.2008.02.003.


Mulcahy, K., Marynluh, M., Peeples, M., Peyrot, M., Tomky, D., & Yarborough, P. (2003).Diabetes self-management education core outcomes. Diabetes Educator, 29(5),768–803, http://dx.doi.org/10.1177/014572170302900509.

Outcome (2014). Consort.com (Retrieved December 12, 2014, from http://www.consort-statement.org/resources/glossary#O).

Pibernik-Okanovic, M., Prasek, M., Poljicanin-Filipovic, T., Pavlic-Renar, I., & Metelko, Z.(2004). Effects of an empowerment-based psychosocial intervention on quality oflife and metabolic control in type 2 diabetic patients. Patient Education andCounseling, 52(2), 193–199, http://dx.doi.org/10.1016/S0738-3991(03)00038-7.

Plummer, M., & Molzahn, A. E. (2009). Quality of life in contemporary nursing theory: Aconcept analysis. Nursing Science Quarterly, 22(2), 134–140, http://dx.doi.org/10.1177/0894318409332807.

Taiwan National Department of Health (2013). Annual medical statistics in Taiwan in 2012.RetrievedMarch 02, 2014, fromhttp://health99.hpa.gov.tw/TXT/HealthyHeadlineZone/HealthyHeadlineDetai.aspx?TopIcNo=6798

TaiwaneseAssociationofDiabetes Educators (2006). The guideline of type 2diabetes in 2006.Retrieved February 20, 2013, from http://www.pntn.doh.gov.tw/main_sec.php?index=hygiene_detail&id=200&no=a12&mid=81&pid=81&sid=04

UK Prospective Diabetes Study Group (1998). Tight blood pressure control and risk ofmacrovascular and microvascular complications in type 2 diabetes: UKPDS 38.British Medical Journal, 317(7160), 703–713.

WorldHealthOrganization (2006).What is the evidence on effectiveness of empowermentto improve health? Retrieved September 02, 2012, from http://www.euro.who.int/__data/assets/pdf_file/0010/74656/E88086.pdf

World Health Organization (2013). Diabetes. Retrieved March 02, 2014, from http://www.who.int/mediacentre/factsheets/fs312/en/

Yip, K. S. (2004). The empowerment model: A critical reflection of empowerment inChinese culture. Social Work, 49(3), 479–487.

Yu, N. C., Su, H. Y., Chiou, S. T., Yeh, M. C., Yeh, S. W., Tzeng, M. S., et al. (2013). Trends ofABC control 2006–2011: A national survey of diabetes health promotion institutes inTaiwan. Diabetes Research and Clinical Practice, 99(2), 112–119, http://dx.doi.org/10.1016/j.diabres.2012.11.018.

Zeger, S. L., & Liang, K. Y. (1986). Longitudinal data analysis for discrete and continuousoutcomes. Biometrics, 42(1), 121–130.












http://dx.doi.org/10.1097/00012272-200109000-00004

http://dx.doi.org/10.1097/00012272-200109000-00004




http://dx.doi.org/10.1891/1541-6577.23.2.107

http://dx.doi.org/10.1891/1541-6577.23.2.107

http://dx.doi.org/10.1177/014572179201800208









http://dx.doi.org/10.1016/j.diabres.2008.02.003


http://dx.doi.org/10.1177/014572170302900509

http://www.consort-statement.org/resources/glossary#O

http://www.consort-statement.org/resources/glossary#O

http://dx.doi.org/10.1016/S0738-3991(03)00038-7

http://dx.doi.org/10.1177/0894318409332807

http://dx.doi.org/10.1177/0894318409332807

http://health99.hpa.gov.tw/TXT/HealthyHeadlineZone/HealthyHeadlineDetai.aspx?TopIcNo=6798

http://health99.hpa.gov.tw/TXT/HealthyHeadlineZone/HealthyHeadlineDetai.aspx?TopIcNo=6798

http://www.pntn.doh.gov.tw/main_sec.php?index=hygiene_detail&id=200&no=a12&mid=81&pid=81&sid=04

http://www.pntn.doh.gov.tw/main_sec.php?index=hygiene_detail&id=200&no=a12&mid=81&pid=81&sid=04




http://www.euro.who.int/__data/assets/pdf_file/0010/74656/E88086.pdf

http://www.euro.who.int/__data/assets/pdf_file/0010/74656/E88086.pdf

http://www.who.int/mediacentre/factsheets/fs312/en/

http://www.who.int/mediacentre/factsheets/fs312/en/








1.efficacy of an empowerment program for taiwanese patients

Documents