Nursing Research

https://doi.org/10.1177/1054773818792480

Clinical Nursing Research 2020, Vol. 29(5) 293 –303 © The Author(s) 2018 Article reuse guidelines: sagepub.com/journals-permissions DOI: 10.1177/1054773818792480 journals.sagepub.com/home/cnr

Research Article

Introduction

Globally, type 2 diabetes mellitus (DM) was the direct cause of an estimated 1.5 million deaths in 2012, whereas hyper- glycemia accounted for another 2.2 million deaths (World Health Organization [WHO], 2016). In Singapore, type 2 DM was the second leading cause of morbidity and mortality in 2015 (Ministry of Health [MOH], 2016). Complications arising from type 2 DM included an increase in new inci- dences of kidney failure from 46% in 1999 to 62% in 2009, with a corresponding increase in end stage renal disease from 28% in 1999 to 44% in 2009 (Singapore Health Promotion Board, 2011). In addition, it was reported that about one in two heart attack cases and two in five stroke cases had type 2 DM in 2014 (MOH, 2016). Type 2 DM was also associated with a 3-fold increase in mortality, of which most were related to cardiovascular diseases, and a 3-fold to 7-fold increase in the risk of coronary artery disease (MOH, 2016). A recent study also revealed that there are now more than 1,500 diabetes-related amputations per year, and one in five

of these patients died within a year of their amputation in Singapore (Ang, Yap, Saxena, Lin, & Heng, 2016).

Background

Given the global shift in aging population, the inevitable demographic change in the prevalence of diabetes and related complications would see a further increase in older adults. The fundamental measure to prevent diabetes-related com- plications is an active lifestyle modification including dietary control, regular physical exercise, medication adherence,

792480CNRXXX10.1177/1054773818792480Clinical Nursing ResearchTan et al. research-article2018

1SingHealth Polyclinics, Singapore 2National University of Singapore, Singapore 3Auckland University of Technology, New Zealand

Corresponding Author: Wenru Wang, Alice Lee Centre for Nursing Studies, Yong Loo Lin School of Medicine, National University of Singapore, Level 2, Clinical Research Centre, Block MD 11,10 Medical Drive, Singapore 117597, Singapore. Email: nurww@nus.edu.sg

Effect of a Diabetes Self-Efficacy Enhancing Program on Older Adults With Type 2 Diabetes: A Randomized Controlled Trial

Cherry Chay Lee Tan, PhD, RN1, Karis Kin Fong Cheng, PhD, RN2, Siew Wai Hwang, MBBS, DR1, Ning Zhang, BScN, RN1, Eleanor Holroyd, PhD, RN3, and Wenru Wang, PhD, RN2

Abstract This randomized controlled trial examined the effect of a diabetes self-efficacy enhancing program (DSEEP) on older adults with type 2 diabetes. The 8-week DSEEP consisted of a guidebook on diabetes self-care, a 1-day workshop, and fortnightly follow-up telephone calls. In total, 113 participants (56 in intervention group and 57 in control group) completed the study. Data were collected at baseline and at 8 weeks from the baseline. Outcome measures included self-efficacy, diabetes self-care activities, health-related quality of life, glycated hemoglobin (HbA1c) and unplanned health care service usage. Compared with participants in the control group, those who received DSEEP had significantly higher increase in self-efficacy and diabetes self-care activities, lower HbA1c, and lesser unplanned health service usage. However, there was no significant difference in health-related quality of life between the two groups. The DSEEP increased self-efficacy, which successfully enhanced self- care activities and reduced HbA1c.

Keywords older adults, chronic illness, clinical effectiveness, type 2 diabetes, self-efficacy, nursing

294 Clinical Nursing Research 29(5)

home self-monitoring of blood glucose, regular foot and eye screening, regular medical follow-up, and stress manage- ment (WHO, 2016). However, studies have reported that age correlates with poorer diabetes self-care management as many old people failed to maintain or achieve blood glucose control (Nguyen et al., 2010). In addition, they are at higher risk for depression and are less likely to engage in self-care activities, leading to an increased risk of complications (Bell et al., 2010). In Singapore, it was predicted that there will be a significant increase in the prevalence of type 2 DM in older adults as the number of Singaporeans aged 65 years and above would triple to 900,000 by 2030 (MOH, 2014). It was also estimated that the number of people with type 2 DM in Singapore will reach 1 million by 2050 (MOH, 2013). These predictions signal an impending health issue that can inflict a considerable health care burden not only on the Singaporean national budget and health care resources but also on indi- vidual patients and their caregivers, families, and communi- ties. Hence, there was an urgent need for new diabetes self-care interventions to address issues faced by older adults and increase their self-care activities.

Diabetes-related complications impose a heavy burden on older adults’ self-care activities, which may affect their self- efficacy expectations. Diabetic neuropathy has often been found upon diagnosis, indicating that older people are at risk of undertreated pain with an increased risk of pressure ulcers, falls, and fractures (Abdelhafiz & Sinclair, 2015; X. Zhang et al., 2010). Many studies also reported that older adults had reduced hypoglycemic awareness because of poorly con- trolled blood glucose over time, and this reduced awareness predisposed older adults to other complications such as diz- ziness and falls (Abdelhafiz, Bailey, Loo, & Sinclair, 2013; Amiel, Dixon, Mann, & Jameson, 2008; Gates & Walker, 2014). Inevitably, these complications adversely affect older adults’ self-efficacy in performing complex diabetes self- care activities such as administering their daily insulin injec- tions, adhering to medicine regime, and performing home self-monitoring blood glucose levels. Bandura (1997) has suggested that older adults need to relook at their self-confi- dence in activities of daily living due to reduced age-related physical, memory, and intellectual capacities.

Self-efficacy, a key construct in the social cognition the- ory, is defined as one’s judgment of his or her capabilities to organize and execute courses of action (Bandura, 2000). It is a determinant of performance, and it is based on the belief of what people think, believe, and feel which affect the way they behave (Bandura, 2000). Indeed, the four informational source of self-efficacy namely performance accomplish- ment, verbal persuasion, vicarious experience, and physio- logical states were built-in strategies to develop behavioral change programs, and studies on their effectiveness have shown positive outcomes in self-care management of people with diabetes. One quasi-experimental study in China tested the effectiveness of a 12-week educational self-efficacy intervention on osteoporosis prevention and diabetes

self-management on older adults (Ha, Hu, Petrini, & McCoy, 2014). The intervention comprised an educational booklet on healthy lifestyle and diabetes self-care; a weekly 1-hr group session of presentations, demonstrations, and discussions; and biweekly telephone calls, and the study results showed significant improvement in diabetes self-efficacy, diabetes self-care activities, and glycaemic control at 3 month’s fol- low-up (Ha et al., 2014). However, few studies focused on older adults in multiracial and multicultural societies in Asia.

Singapore is a multiracial and multicultural society with three major ethnic groups of Chinese, Malay, and Indian. Prior to this study, four focus group discussions were con- ducted with these three groups to explore the experiences of older adults in diabetes self-care management, and the key issues found were misperceptions about the severity of type 2 DM, deficit in diabetes knowledge, difficulties in diabetes self-care activities, forgetfulness in medication administra- tion, having no time for physical activities, and a lack of self- confidence in diabetes self-care skills (Tan et al., 2018). Hence, this study aimed to examine the effect of a newly developed ethnic sensitive diabetes self-efficacy enhancing program (DSEEP) on self-efficacy, self-care activities, Health-Related Quality of Life (HRQoL), glycated hemoglo- bin (HbA1c), and unplanned health care service usage among older adults with type 2 diabetes in Singapore. It was hypoth- esized that compared with the control group, the intervention group will have statistically significant increases in self-effi- cacy and self-care activities, improved HRQoL, and reduced HbA1c and unplanned health care service usage after the intervention.

Method

Design and Setting

A randomized controlled trial with a pre- and posttest control group design was adopted. The study was conducted in a polyclinic (i.e., primary care setting) in Singapore, where approximately 2,000 patients with diabetes follow up regu- larly with the doctors at an interval between 3 and 4 months to monitor their diabetic condition. The first author briefed all the doctors and nurse counselors about the study and advised them to refer prospective participants to the first author for enrollment.

Sample

A consecutive sampling method was adopted, and the data were collected from June 2015 to June 2016. The inclusion criteria were patients who had a confirmed clinical diagnosis of type 2 DM were 50 years old and above, had HbA1c of >8% in their most recent test, and were able to read and speak English or Chinese. Those patients who had severe stroke, visual impairment, renal failure, history of major psychiatric illness, and/or major hearing difficulties were excluded.

Tan et al. 295

Previous studies investigating behavioral change showed that self-efficacy for diabetes self-care management in their experimental groups was significantly higher with a medium effect size (Hawkins, 2010; Wu et al., 2011). A sample size of 64 in each group can detect differences between the two groups at an alpha of .05 and a power of 0.80 (Machin, Campbell, Fayers, & Pinol, 1997). The total sample size was 142 older adults, including an estimated attrition rate of 10% based on studies of diabetes self-care management that var- ied from 0.8% to 16% (Hawkins, 2010; Wu et al., 2011).

Participants’ Recruitment and Randomization

The participants were recruited during their visit at the poly- clinic for regular follow-ups or medical consultations. Nurse counselors referred patients who met the inclusion criteria to the first author who then assessed the participants’ eligibility for enrollment and explained all aspects of the study includ- ing the study aim, study procedure, and their rights to with- draw at any time during the course of the study by informing the first author. The participant gave written informed con- sent after they agreed to participate. The first author allo- cated the participants to either the intervention group (A) or the control group (B) using blocked randomization with six combinations for block size 4: AABB, ABAB, ABBA, BBAA, BABA, and BAAB.

The Trial Interventions

Usual care. Both the intervention and control groups contin- ued to receive the usual care provided by the polyclinic. Usual care includes follow-up consultation and treatment with their doctors, HbA1c monitoring, diabetic retinopathy photography, diabetic foot screening, and face-to-face coun- seling with a nurse educator on self-care management such as eating healthily and engaging in regular physical activities.

Diabetes self-efficacy enhancing program (DSEEP). The DSEEP is an 8-week program comprising a DM guidebook, supple- mented with a digital videodisc, a 1-day training workshop, and three fortnightly follow-up telephone calls. The develop- ment of the 8-week DSEEP was guided by Bandura’s (1997) self-efficacy theory. Self-efficacy is deemed an ever chang- ing construct as its judgment changes over time with acquisi- tion of new information that could come from four sources, namely performance accomplishment, vicarious experience, verbal persuasion, and physiological feedback (Resnick, 2004). Accordingly, the DSEEP focused on inducing partici- pants’ self-efficacy expectations through the acquisition of DM knowledge and self-care skills from the four sources of information. The intervention group was expected to have an increased self-efficacy through increased diabetes knowl- edge from a guidebook and an accompanied DVD on living life with diabetes, mastered diabetes self-care skills during a

1-day training workshop, and exposure to learning strategies during the follow-up telephone calls, whereby advices were provided to overcome any difficulties faced in self-care activities as indicated in the theoretical framework (Figure 1). The intervention group underwent DSEEP and usual care, whereas the control group received only the usual care. The first author conducted the DSEEP to ensure consistency in the delivery of the intervention for all groups. A pilot study with a total of 42 participants was conducted to test the fea- sibility of the intervention prior to this study (Mitsui Sumi- tomo Insurance Group, 2018). Although the statistical results from the pilot study were not significant due to the small sample size, it showed that the intervention group as com- pared with the control group had higher increase in self-effi- cacy and self-care activities and a higher reduction in HbA1c. Hence, the intervention was found feasible, and the data col- lected were included in this study.

Data Collection

Data with questionnaires were collected by the first author in person with participants from both the control and the inter- vention groups prior to the intervention and 8 weeks imme- diately after the intervention. This is to ensure that all clarifications if needed from participants were answered consistently. The participants’ clinical data were obtained from their medical record.

Primary outcome. The primary outcome was self-efficacy, and this was measured using the General Self-efficacy Scale (GSS). The GSS consists of 10 items with four possible responses, ranging from “absolutely incorrect” to “abso- lutely correct.” Response to all 10 items produces a total score ranging from 10 to 40, with a higher score indicating higher positively perceived self-efficacy (Schwarzer & Jeru- salem, 1995; J. X. Zhang & Schwarzer, 1995).

Secondary outcomes. The Revised Summary of Diabetes Self-care Activities (RSDSA) scale was used to assess par- ticipant’s diabetes self-care activities. The RSDSA com- prises 10 items on five domains of self-care activities, namely diet, exercise, medication, foot care, and blood glucose mon- itoring. Each domain consists of two items, ranged from “0” to “7,” day on which activities were performed weekly. A higher score indicates a higher number of self-care activities performed (Toobert, Hampson, & Glasgow, 2000; Xu, Sav- age, Toobert, Pan, & Whitmer, 2008).

The participant’s HRQoL was assessed using the Audit of Diabetes-Dependent Quality of Life (ADDQoL). The scale comprises two overview items (i.e., HRQoL-General) and 19 domain-specific questions to rate the impact of dia- betes on various aspects of living including leisure, work, travel, physical activities, family life, social life, personal life, sex life, self-confidence, financial, and living condi- tions (i.e., HRQoL-19-Domain). Each overview item is

296 Clinical Nursing Research 29(5)

scored individually, whereas scoring for each individual domain utilizes a weighted score that was calculated by multiplying the impact rating (–3 to +1) with the impor- tance rating (0-3). The weighted impact score eventually produces a total score that ranges from −9 to +3. A nega- tive score indicates a negative impact of diabetes, and a positive score indicates a positive impact of diabetes (Soon et al., 2010; Wee, Tan, Goh, & Lee, 2006).

HbA1c values were retrieved from the clinic’s laboratory. Questionnaires were specifically designed for the collection of demographic/clinical data and unplanned health care ser- vice usage data, which include hospital admission, visit to the accident and emergency department, and medical consul- tations that were diabetes related.

Validity and Reliability

In this study, randomization minimized confounding fac- tors including social background, age, education qualifica- tions, and so forth. All the participants had an equal chance of allocation into the intervention or the control group. Indeed, the study results showed no significant differences in demographic and clinical characteristics between the two groups at baseline. In addition, all data were collected through face to face with the participants at baseline and postintervention time points to enhance the response rate, and the opportunity to better clarify questions for the par- ticipants during data collection.

The instruments used in this study have been well vali- dated, and this ensured the trustworthiness of the results. The English version of the GSS (Cronbach’s α = .76 to .90), and

its Chinese version (Cronbach’s α = .91) for the total scale showed good internal consistency and test–retest reliability (Schwarzer & Jerusalem, 1995; J. X. Zhang & Schwarzer, 1995). The English version of the RSDSA scale had an aver- age interitem correlation within scales (M = 0.47) and a moderate test–retest correlation (M = 0.40; Toobert et al., 2000), whereas its Chinese version (Cronbach’s α = .62 to .87) was validated in the Chinese population (Xu et al., 2008). The English version of the ADDQoL (Cronbach’s α = .94) and its Chinese version (Cronbach’s α = .94) were examined and validated locally (Soon et al., 2010; Wee et al., 2006). In our study, the GSS, RSDSA, and ADDQoL have also shown good reliability with Cronbach’s alpha of .81, .72 and .70, respectively.

Ethical Consideration

Ethical approval was obtained from the Domain-Specific Review Board of the National Health care Group in Singapore in March 2015 (NHG DSRB reference number: 2015/00042).

Data Analysis

The Statistical Package for the Social Sciences (SPSS) for Windows (22.0, SPSS Institute, Chicago, IL) was used. Descriptive statistics described the samples, and data were summarized from dependent variables. Homogeneity (sociodemographic and clinical variables) of groups for nor- mally distributed continuous variables was determined using independent t tests, whereas Chi-square tests were used for nominal variables. Paired t test was used to test the difference

Figure 1. Underpinning theoretical framework. Note. DSEEP = diabetes self-efficacy enhancing program; HRQoL = Health-Related Quality of Life; HbA1c = glycosylated hemoglobin.

Tan et al. 297

of mean scores of study outcomes within each group, whereas repeated measures ANOVA was performed to test the interac- tion effect (group × time) between the two groups across 2-month study period. Significance level was set at 0.05.

For normal distribution of data, Shapiro–Wilk Tests were carried out using the baseline scores of outcome vari- ables for both groups examined. In cases of skewness in distribution, the QQ plots further examined the feasibility for normality assumption. For missing data that arose from participants who did not complete the study protocol, a remediation approach was conducted to ascertain differ- ences in demographic characteristics and baseline data between the participants who completed the study protocol and those who violated it.

Results

Characteristics of participants. There were about 185 prospec- tive participants referred by the health counselors for

eligibility assessment to participate in this study between June 2015 and January 2016. Out of the 185 referred, 142 were recruited and randomized to either the intervention or the control groups. Twenty-nine participants further dropped out due to work commitment, overseas travel, or personal reasons. Out of this, 20 dropped out before the start of the intervention, whereas nine withdrew after baseline data col- lection. The 113 participants (56 in intervention and 57 in control) who completed the whole study were included in the data analysis (Figure 2). A remediation of the missing data showed no significant differences between the two groups in baseline data, implying that the factors that cause the missing data were unrelated to the intervention effect.

In terms of sociodemographic and clinical characteristics of the participants, the mean age of participants in the inter- vention and control groups was 61.50 (SD = 6.78) years and 62.75 (SD = 7.20) years, respectively; the mean length of diagnosis for type 2 DM was 11.3 (SD = 6.8) years and 13.4 (SD = 9.2) years, respectively, and HbA1c was 9.86%

Figure 2. The CONSORT flow diagram for data collection. Note. DSEEP = diabetes self-efficacy enhancing program; GSS = General Self-efficacy Scale; RSDSA = Revised Summary of Diabetes Self-Care Activities; ADDQoL = Audit of Diabetes-Dependent Quality of Life; UHSU = unplanned health service usage.

298 Clinical Nursing Research 29(5)

(SD = 1.7%) and 9.58% (SD = 1.3%), respectively. There were no significant differences in terms of sociodemographic and clinical characteristics as well as study outcomes between the two groups at the baseline.

Self-Efficacy Findings

The mean scores of GSS increased from the baseline to the posttest in both intervention and control groups (Table 1). However, paired t test indicated that significant difference of mean scores of GSS was found only in the intervention group (t = 3.35, p = .001) but not in the control group (t = 0.25, p = .80). A repeated measures ANOVA showed a significant interaction effect (group × time) on GSS between the two groups, F(1, 111) = 4.99, p = .03 (Figure 3a), indicating that the trend of GSS means in the intervention group signifi- cantly increased compared with the control group across the 2-month study period.

Diabetes Self-Care Activities Findings

The mean score of RSDSA also increased from the baseline to the posttest in both intervention and control groups. However, the significant increase of RSDSA mean was only found in the intervention group (t = 7.45, p < .001) but not in the control group (t = 2.06, p = .05; Table 1). A repeated measures ANOVA showed a significant interaction effect (group × time) on RSDSA between the two groups, F(1, 111) = 18.11, p < .01 (Figure 3b), indicating the trend of RSDSA means in the intervention group significantly increased

compared with the control group across the 2-month study period.

Health-Related Quality of Life Findings

HRQoL-General. The negative mean score for HRQoL- General decreased significantly for the intervention group (t = −2.79, p = .009), indicating that the negative impact of DM on the participants’ HRQoL-General reduced significantly after the intervention. The negative impact of DM on participants’ HRQoL-General in the control group also reduced, but it was not significant (t = −0.91, p = .40). The repeated measures ANOVA showed no significant interaction effect (group × time) between the two groups on HRQoL-General, F(1, 111) = 1.37, p = .24 (Figure 3c), indicating no difference of mean scores for HRQoL-General across 2-month study period between the two groups.

HRQoL-19-Domain. The negative mean score for HRQoL- 19-Domain increased for the intervention group but it was not significant (t = −1.38, p = .17). Conversely, the negative mean score for the control group decreased significantly (t = −2.11, p = .04), indicating that the negative impact of DM on these participants’ HRQOL-19-Domain improved (Table 1). The repeated measures ANOVA showed no sig- nificant interaction effect (group × time) between the two groups on HRQoL (19-Domain), F(1, 111) = 6.11, p = .05 (Figure 3d), indicating no differences of mean scores for HRQoL-19-Domin across 2-month study period between the two groups.

Table 1. Outcome Variables at the Baseline and Posttest Within and Between the Two Groups.

Outcome variables

Baseline Posttest Changes of mean between baseline

and posttest t p

Interaction effect

M (SD) M (SD) F P

GSS 4.99 .03* Intervention group (n = 56) 3.02 (0.62) 3.29 (0.50) 0.27 3.35 .001** Control group (n = 57) 2.83 (0.62) 2.85 (0.52) 0.02 0.25 .80 RSDSA 18.11 <.001** Intervention group (n = 56) 3.63 (1.21) 4.81 (1.11) 1.18 7.45 <.001** Control group (n = 57) 3.64 (1.44) 3.92 (1.35) 0.28 2.06 .05 HRQoL-General 1.37 .24 Intervention group (n = 56) −0.32 (0.70) −0.05 (0.49) 0.27 −2.79 .009* Control group (n = 57) −0.46 (0.71) −0.36 (0.61) 0.10 −0.91 .40 HRQoL-19-Domain 6.11 .05 Intervention group (n = 56) −3.79 (2.23) −4.06 (2.12) −0.27 −1.38 .17 Control group (n = 57) −3.72 (2.24) −3.30 (2.18) −0.42 −2.11 .04* HbA1c 6.08 .01* Intervention group (n = 56) 9.90 (1.72) 8.66 (1.24) −1.24 −5.53 <.001** Control group (n = 57) 9.58 (1.38) 9.04 (1.54) −0.58 −3.41 .001**

Note. T = paired t test; F = repeated measure ANOVA; GSS = General Self-Efficacy Scale; RSDSA = Revised Summary of Diabetes Self-care Activities; HRQoL = Health-Related Quality of Life; HbA1c = glycosylated hemoglobin. *p < .05. **p < .01.

Tan et al. 299

Figure 3. The trend of GSS, DSA, HRQoL, and HbA1c between two groups over time. Note. GSS = General Self-efficacy Scale; HRQoL = Health-Related Quality of Life; HbA1c = glycosylated hemoglobin; DSA = Diabetes Self-care Activities.

300 Clinical Nursing Research 29(5)

Although the findings showed significant improvement in HRQoL-General for the intervention group as compared with the control group, the intervention group showed no improvement in HRQoL-19-Domain. Hence, the difference between the two groups was deemed not significant for HRQoL in totality.

HbA1c Findings

There was a significant reduction of HbA1c from baseline to the posttest in both the intervention group (t = −5.53, p< .001) and the control group (t = −3.41, p< .001; Table 1). However, repeated measures ANOVA showed a significant interaction effect (group × time) between the two groups, F(1, 111) = 6.08, p = .01 (Figure 3e). The intervention group had better glycaemic control with a significant higher reduction in HbA1c from the baseline to the posttest com- pared with the control group across the 2-month study period.

Unplanned Health care Service Usage Findings

There was no report of diabetes-related hospital admission by the participants in both groups at the baseline or at post- test. However, four participants (7.14%) in the intervention group reported diabetes-related unplanned medical consulta- tion at the baseline, whereas no patient reported unplanned medical consultation at posttest. For the control group, three participants (5.26%) reported diabetes-related unplanned medical consultation at the baseline, whereas one patients (1.75%) had unplanned medical consultation at posttest. This indicates that the intervention group had a higher reduction in frequency of diabetes-related unplanned medical consulta- tion (–7.14%) compared with the control group (–3.51%). The control group also had new cases of diabetes-related vis- its to the accident and emergency department (n = 2, 3.51%) whereas there was no visit for the intervention group during the 2-month study period.

Discussion

This study results show that the participants who underwent DSEEP had significant increase in self-efficacy, diabetes self-care activities, HbA1c, and lesser unplanned health care service usage compared with the control group across the 2-month study period. For HRQoL in general, the interven- tion group also had significant increase but not for HRQoL- 19-Domain. Hence, the hypotheses that participants in the intervention group compared with the control group will have significant increase in self-efficacy, diabetes self-care activities, and HbA1c were accepted.

The increased self-efficacy is likely to have resulted in increase in diabetes self-care activities as participants became more confident in executing their daily self-care with more diabetes knowledge and improved self-care skills gained from the training workshops. Indeed, the findings on

self-efficacy for the intervention group is consistent with previous studies that reported on various health behavior changes and adaptations, which were achieved by interven- tions that focused on enhancing self-efficacy in self-care management (Hernandez et al., 2016; Nelson, McFarland, & Reiber, 2007; Trief, Teresi, Eimicke, Shea, & Weinstock, 2009; Wu et al., 2011). Enhanced self-efficacy motivated health behavioral change has been associated with increased diabetes self-care activities, leading to reduced health care service usage (Hernandez et al., 2016; Nelson et al., 2007; Trief et al., 2009; Wu et al., 2011). Conversely, poorer self- efficacy in older adults resulted in lack of motivation to change health behaviors, as well as nonadherence to medica- tion (Bean, Cundy, & Petrie, 2007; Nelson et al., 2007).

The increase in diabetes self-care activities for the inter- vention group is also similar to the findings of many studies, which evaluated the effects of interventions that focused on self-efficacy and its association with self-care activities (Ha et al., 2014; Hawkins, 2010; Hernandez et al., 2016; Wu et al., 2011). Conversely, studies have also found that self- efficacy rate was low among those with poor self-care activi- ties, and a high level of self-efficacy was associated with a high level of self-care activities (Beckerle & Lavin, 2013; Redmond et al., 2006). It was suggested that there was a direct relationship between self-efficacy and self-care, whereby the construct of self-efficacy highly predicts self- care behavior (Gao et al., 2013; Mohebi, Azadbakht, Feizi, Sharifirad, & Kargar, 2013). In this study, the significant increase in diabetes self-care activities could also be due to the counseling and advice given to the participants in the intervention group during the follow-up telephone calls post intervention. It can be postulated that participants were likely to have followed the nurse researcher’s advice to overcome any difficulties encountered in self-care activities. This cor- responds with previous studies that found positive outcomes in diabetes self-care activities by increasing patients’ knowl- edge via telemedicine (Berg & Wadhwa, 2007; Wangnoo et al., 2013). Indeed, the increase in diabetes self-care activi- ties have led to reduced diabetes-related complications as evidenced from the reduction in diabetes-related unplanned health care service usage for the intervention group.

Findings for HRQoL-General indicated that the DSEEP had generally improved the quality of life of the participants in the intervention group but showed an increase in negative effect for their HRQoL-19-Domain. This contradicts studies that reported improved diabetes self-care activities with improved quality of life (Huang & Hung, 2007; Mohebi et al., 2013). A further analysis of the individual domains in the HRQoL-19-Domain scale for the intervention group showed a significant increase in negative mean scores for five domains including feeling about the future, freedom to eat, freedom to drink, depending on others, and motivation. The possible underlying reason for this contradictory effect of the HRQoL-19-Domain could be the short duration of exposure to the intervention. The 8-week exposure may not

Tan et al. 301

be sufficient for participants to adjust to the increased self- care activities at the initial stage. Future study may be needed to measure HRQoL for older adults with a longer duration of exposure to the intervention.

Both the intervention and the control groups had signifi- cant reduction in HbA1c. The improvement in the control group is likely due to usual care, whereby drug therapy is usually adjusted to achieve the optimal HbA1c target. However, the intervention group had a greater improvement in HbA1c than the control group, and this positive results of the DSEEP on older adults’ HbA1c is consistent with many studies where an increase in self-efficacy was associated with an increase in diabetes self-care activities, which in turn resulted in reduced HbA1c and better glycaemic control (Gao et al., 2013; Hawkins, 2010; Silva, Clinton, Appleton, & Flanagan, 2011). This finding also indicates that self-effi- cacy played a key role in mediating diabetes self-care activi- ties, which in turn resulted in the significantly large reduction in HbA1c of 1.24% for the intervention group versus 0.58% for the control group. A reduction of HbA1c by just 1% can reduce the risk of diabetes-related complications including diabetes-related mortality by 21%, myocardial infarction by 14%, and microvascular complications by 37% (International Diabetes Federation, 2016). The positive outcome of HbA1c showed that the DSEEP is not only effective in achieving good glycaemic control, it can also help older adults reduce the risks of lifelong complications. Correspondingly, the fre- quency of unplanned health care service usage was also lesser for the intervention group with a higher reduction in diabetes-related unscheduled medical consultation. However, the control group had increased visits to the accident and emergency department due to diabetes-related complica- tions. It also implies that the intervention group had lesser diabetes-related complications after the DSEEP. This is also consistent with a previous study on the effect of an educa- tional health care package that reported significantly fewer complications in older adults from the intervention group at 6 months follow-up when compared with the control group (Ha et al., 2014).

Limitations

The study has several limitations. There was no blinding strategy to prevent the group allocator from knowing the intervention received by the participants. As this study is part of an academic pursuit for the first author who had to con- duct the study herself, it was not possible to blind the first author to the randomization process. Concealed allocation was also impossible because the intervention group needed to undergo the DSEEP and cannot be blinded to the interven- tion. The use of self-reported questionnaires for measuring the outcomes was subjected to social desirability and recall bias. In addition, the high attrition rate is also a threat to external validity, though a remediation of the missing data showed no significant difference between those who dropped

out from the study and those who completed the study. However, the HbA1c as an objective measure revealed that the DSEEP was effective in glycaemic control, and self-effi- cacy is a likely mediator given that the participants who underwent the program showed better improvement in HbA1c. The long-term effect of DSEEP is lacking in this study. Future research could possibly examine its long-term effect on older adults with type 2 DM.

Generalizability of Trial Findings

In Singapore, the Chinese (74.3%) is the dominant ethnic group, and the Malay (13.4%) and Indian (9.0%) are the minority ethnic groups in 2012 (Singapore Department of Statistics, 2017). In terms of diabetes prevalence, the Indians (17.2%) had the highest percentage of diabetes compared with the Malay (16.6%) and Chinese (9.7%; MOH, 2013). In this study, the Chinese (68.1%), Indian (20.4%), and Malay (11.5%) participants represented the three ethnicities propor- tionately in terms of their population including diabetes preva- lence, and this enhanced the generalizability of the study outcomes.

Conclusion

The positive outcomes from this study showed that diabetes self-care intervention guided by the self-efficacy theory can impact on older adults’ self-care management, leading to bet- ter glycaemic control and lower risks of diabetes-related complications. In addition, the DSEEP was effective in older Singaporean adults in enabling their diabetes self-care man- agement with the findings which supported the hypotheses that participants in the intervention group compared with those in the control group had significantly improved self- efficacy, increased diabetes self-care activities, and reduced HbA1c. The DSEEP was also deemed effective in reducing unplanned health care service usage. Therefore, the DSEEP could be a new option for health care service delivery to improve the self-care management of older adults with type 2 DM in Singapore.

Relevance to Clinical Practice

The DSEEP could be an alternative intervention for nursing management to meet the needs of older adults with type 2 DM in their self-care management. Notably, it has been rec- ognized that the primary resource for diabetes self-care man- agement is the patients themselves. Hence, diabetes nurse educators and clinicians should focus on empowering older adults with knowledge and skill to take charge of their diabe- tes management. This, in turn, enables older adults to assume the responsibilities and accountability for the control of their own condition. In addition, the delivery of DSEEP could also be an alternative method to enhance the self-efficacy of older adults to increase diabetes self-care activities at home.

302 Clinical Nursing Research 29(5)

Numerous studies had reported positive outcomes from interventions that enhance the self-efficacy of patients with type 2 DM, and this study had also demonstrated the value of applying DSEEP to older adults in Singapore. Older adults can carry out diabetes self-care activities effectively with the guide to living life with diabetes at the comfort of their home. The guide can also be offered to their caregivers as a resource to understand the condition of their loved ones and to use it as a guide in their care for them.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the Mitsui Sumitomo Insurance Welfare Foundation Research Grant for Senior Citizen Welfare.

ORCID iD

Wenru Wang https://orcid.org/0000-0002-0265-8215

References

Abdelhafiz, A. H., Bailey, C., Loo, B. E., & Sinclair, A. (2013). Hypoglycemic symptoms and hypoglycemia threshold in older people with diabetes—a patient perspective. Journal of Nutrition Health & Aging, 17, 899-902.

Abdelhafiz, A. H., & Sinclair, A. J. (2015). Epidemiology, aetiology, pathogenesis and management of diabetes in older people. In L. Goff & P. Dyson (Eds.), Advanced nutrition and dietetics in dia- betes (pp. 207-217). Retrieved from http://onlinelibrary.wiley. com.libproxy1.nus.edu.sg/book/10.1002/9781119121725

Amiel, S. A., Dixon, T., Mann, R., & Jameson, K. (2008). Hypoglycaemia in type 2 diabetes. Diabetes Medicine, 25, 245-254. doi:10.1111/j.1464-5491.2007.02341.x

Ang, Y., Yap, C. W., Saxena, N., Lin, L.-K., & Heng, B. H. (2016). Diabetes-related lower extremity amputations in Singapore. Proceedings of Singapore Healthcare, 26, 76-80. doi:10.1177/2010105816663521

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

Bandura, A. (2000). Health promotion from the perspective of social cognition theory. In P. Norman, C. Abraham, & M. Conner (Eds.), Understanding and changing health behaviour: From health beliefs to self-regulation (pp. 299-339). Singapore: Harwood academic publishers.

Bean, D., Cundy, T., & Petrie, K. J. (2007). Ethnic differences in illness perceptions, self-efficacy and diabetes self-care. Psychology & Health, 22, 787-811.

Beckerle, C. M., & Lavin, M. A. (2013). Association of self- efficacy and self-care with glycemic control in diabetes. Diabetes Spectrum, 26, 172-178.

Bell, R. A., Andrews, J. S., Arcury, T. A., Snively, B. M., Golden, S. L., & Qiandt, S. A. (2010). Depressive symptoms and dia- betes self-management among rural older adults. American Journal of Health Behavior, 34(1), 36-44.

Berg, G. D., & Wadhwa, S. (2007). Health services outcomes for a diabetes disease management program for the elderly. Disease Management, 10, 226-234.

Gao, J. J., Wang, J., Zheng, P., Haardörfer, R., Kegler, M. C., Zhu, Y., & Fu, H. (2013). Effects of self-care, self-efficacy, social support on glycemic control in adults with type 2 diabetes. BMC Family Practice, 14, Article 66. doi:10.1186/1471-2296-14-66

Gates, B. J., & Walker, K. M. (2014). Physiological changes in older adults and their effect on diabetes treatment. Diabetes Spectrum, 27(1), 20-29.

Ha, M., Hu, J., Petrini, M. A., & McCoy, T. P. (2014). The Effects of an educational self-efficacy intervention on osteoporosis prevention and diabetes self-management among adults with type 2 diabetes mellitus. Biological Research for Nursing, 16, 357-367. doi:10.1177/1099800413512019

Hawkins, S. Y. (2010). Improving glycemic control in older adults using a videophone motivational diabetes self-management intervention. Research and Theory for Nursing Practice, 24, 217-232.

Hernandez, R., Ruggiero, L., Prohaska, T. R., Chavez, N., Boughton, S. W., Peacock, N., . . . Nouwen, A. (2016). A cross-sectional study of depressive symptoms and diabetes self-care in African Americans and Hispanics/Latinos with diabetes. The role of self-efficacy. The Diabetes Educator, 42, 452-461. doi:10.1177/0145721716654008

Huang, M., & Hung, C. (2007). Quality of life and its predictors for middle-aged and elderly patients with type 2 diabetes mellitus. The Journal of Nursing Research, 15, 193-201.

International Diabetes Federation. (2016). Fact sheet 1: Good gly- cemic control. Retrieved from http://www.fend.org/sites/fend. org/files/CTG_Fact_sheets.pdf

Machin, D., Campbell, M., Fayers, P., & Pinol, A. (1997). Sample size tables for clinical studies. Oxford, UK: Blackwell Science.

Ministry of Health Singapore. (2013). National Health Survey 2010. Diabetes mellitus. Retrieved from https://www.moh. gov.sg/content/moh_web/home/Publications/Reports/2011/ national_health_survey2010.html

Ministry of Health. (2014). Healthy living master plan. Retrieved from https://www.moh.gov.sg/content/dam/moh_web/Publica tions/Reports/2014/HLMP/MOH_Healthy%20Living%20 Master%20Plan_Inside%20Page_8d.pdf

Ministry of Health. (2016). Better health, better care, better life. The war on diabetes. Rise of diabetes in Singapore. Retrieved from https://www.moh.gov.sg/content/dam/moh_web/PressRoom/ Highlights/2016/cos/factsheets/COS_Factsheet%20-%20 Diabetes.pdf

Mitsui Sumitomo Insurance Group. (2018). Mitsui sumitomo insur- ance welfare foundation research grant 2014. Retrieved from https://imsva91-ctp.trendmicro.com:443/wis/clicktime/v1/ query?url=http%3a%2f%2fwww.ms%2dins.com%2fwelfa re%2fdocument%2flist%2fpdf%2f2014%2f3%5f1%5f27. pdf&umid=509641F0-66DC-8B05-919D-56026F025DE 7&auth=6e3fe59570831a389716849e93b5d483c90c3fe4- ed39b275ac5c8559a39d5eba9d0fbb6cd4061e55

Mohebi, S., Azadbakht, L., Feizi, A., Sharifirad, G., & Kargar, M. (2013). Review the key role of self-efficacy in diabetes care. Journal of Education and Health Promotion, 2, Article 36. doi:10.4103/2277-9531.115827

Nelson, K. M., McFarland, L., & Reiber, G. (2007). Factors influ- encing disease self-management among veterans with diabetes

Tan et al. 303

and poor glycemic control. Journal of General Internal Medicine, 22, 442-447.

Nguyen, H. T., Grzywacz, J. G., Arcury, T. A., Chapman, C., Kirk, J. K., Ip, E. H., . . . Quandt, S. A. (2010). Linking glycemic control and executive function in rural older adults with diabe- tes mellitus. Journal of the American Geriatrics Society, 58, 1123-1127.

Redmond, E. H., Burnett, S. M., Johnson, M. A., Park, S., Fischer, J. G., & Johnson, T. (2006). Improvement in A1C levels and diabetes self-management activities following a nutrition and diabetes education program in older adults. Journal of Nutrition for Elderly, 26, 83-102.

Resnick, B. (2004). Self-efficacy. In S. J. Peterson & T. S. Bredow (Eds.), Middle range theories. Application to nursing research (pp. 117-146). Philadelphia, PA: Lippincott William & Wilkins.

Schwarzer, R., & Jerusalem, M. (1995). The Generalized Self- efficacy Scale. In J. Weinman, S. Wright, & M. Johnston (Eds.), Measures in health psychology: A user’s portfolio. Causal and control beliefs (pp. 35-37). Retrieved from https:// userpage.fu-berlin.de/health/engscal.htm

Silva, M., Clinton, J., Appleton, S., & Flanagan, P. (2011). Diabetes self-management education in South Auckland, New Zealand, 2007-2008. Preventing Chronic Disease, 8, A42.

Singapore Department of Statistics. (2017). Population Trend 2017. Retrieved from https://www.singstat.gov.sg/-/media/ files/publications/population/population2017.pdf

Singapore Health Promotion Board. (2011). Information paper on diabetes in Singapore. Available from http://www.hpb.gov.sg

Soon, S. S., Goh, S. Y., Bee, Y. M., Poon, J. L., Li, S. C., Thumboo, J., & Wee, H. L. (2010). Audit of Diabetes-Dependent Quality of Life (ADQoL) [Chinese version for Singapore] questionnaires: Reliability and validity among Singaporeans with type 2 diabe- tes mellitus. Applied Health Economics and Health Policy, 8, 239-249.

Tan, C. C. L., Cheng, K. K. F., Sum, C. F., Shew, J. S. H., Holydard, E., & Wang, W. (2018). Perceptions of diabetes self-care man- agement among older Singaporeans with type 2 diabetes: A qualitative study. The Journal of Nursing Research, 26, 242- 249. doi:10.1097/jnr.0000000000000226

Toobert, D. J., Hampson, S. E., & Glasgow, R. E. (2000). The sum- mary of diabetes self-care activities measure: Results from 7 studies and a revised scale. Diabetes Care, 23, 943-950.

Trief, P. M., Teresi, J. A., Eimicke, J. P., Shea, S., & Weinstock, R. S. (2009). Improvement in diabetes self-efficacy and glycae- mic control using telemedicine in a sample of older, ethnically diverse individuals who have diabetes: The IDEATel project. Age and Ageing, 38, 219-225. doi:10.1093/ageing/afn299

Wangnoo, S. K., Maji, D., Das, A. K., Rao, P. V., Moses, A., Sethi, B., . . . Dharmalingam, M. (2013). Barriers and solutions to diabetes management: An Indian perspective. Indian Journal of Endocrinology & Metabolism, 17, 594-601.

Wee, H. L., Tan, C. E., Goh, S. Y., & Lee, S. E. (2006). Usefulness of the Audit of Diabetes-Dependent Quality-of-Life (ADDQoL)

questionnaire in patients with diabetes in a multi-ethnic Asian country. PharmacoEconomic, 24, 673-682.

World Health Organisation. (2016). Diabetes: Global report on diabetes. Retrieved from http://www.who.int/diabetes/global- report/en/

Wu, S. F. V., Lee, M. C., Liang, S. Y., Lu, Y. Y., Wang, T. J., & Tung, H. H. (2011). Effectiveness of a self-efficacy pro- gram for persons with diabetes: A randomized controlled trial. Nursing & Health Sciences, 13, 335-343. doi:10.1111/j.1442- 2018.2011.00625.x

Xu, Y., Savage, C., Toobert, D., Pan, W., & Whitmer, K. (2008). Adaptation and testing of instruments to measure diabetes self- management in people with type 2 diabetes in mainland China. Journal of Transcultural Nursing, 19, 234-242.

Zhang, J. X., & Schwarzer, R. (1995). Measuring optimistic self- beliefs: A Chinese adaptation of the general self-efficacy scale. Psychologia, 38, 174-181.

Zhang, X., Decker, F. H., Luo, H., Geiss, L. S., Pearson, W. S., Saaddine, J. B., . . . Albright, A. (2010). Trends in the preva- lence and comorbidities of diabetes mellitus in nursing home residents in the United States: 1995–2004. Journal of the American Geriatrics Society, 58, 724-730.

Author Biographies

Cherry Chay Lee Tan, PhD, RN, is a senior staff nurse working in SingHealth Polyclinics in Singapore. Her main research interests include patient education, nursing care of patients with diabetes.

Karis Kin Fong Cheng, PhD, RN, is a full professor at Alice Lee Centre for Nursing Studies, Yong Loo Lin School of Medicine, National University of Singapore. Professor Cheng’s research inter- ests include cancer therapy-related complication and side effect management, and the quality of life and psychometric evaluation.

Siew Wai Hwang, MBBS, DR, is the clinical director of SingHealth Polyclinics in Singapore. He has extensive clinical working experi- ences for patients with diabetes.

Ning Zhang, BScN, RN, is a senior staff at SingHealth Polyclinics in Singapore. She has extensive clinical experiences taking care of patients with diabetes.

Eleanor Holroyd, PhD, RN, is a full professor of nursing at Auckland University of Technology, New Zealand. Professor Holroyd’s research interests are in ethnography, sexual and repro- ductive health (in particular HIV research), migrant research, female sex-worker health anthropology and women’s health.

Wenru Wang, PhD, RN, is a tenured associate professor at Alice Lee Centre for Nursing Studies, Yong Loo Lin School of Medicine, National University of Singapore. Her major research interests include cardiovascular nursing and cardiac rehabilitation, develop- ing psychometric measurement tools, chronic diseases/symptoms management, and nursing education.