Collegian (2014) 21, 4351

Available online at www.sciencedirect.com

j our nal homepage: www. el sevi er . com/ l ocat e/ col l
Cluster randomised controlled trial: Educational
self-care intervention with older Taiwanese patients
with Type 2 diabetes mellitusImpact on blood
glucose levels and diabetic complications
Ying-Hua Chao
a,b
, Kim Usher
b,
, Petra G. Buettner
c
, Colin Holmes
d
a
Department of Nursing, Yuanpei University, Taiwan
b
School of Nursing, Midwifery and Nutrition, James Cook University, Cairns, PO Box 6811, Cairns, QLD 4870, Australia
c
School of Public Health, Tropical Medicine and Rehabilitation Sciences, James Cook University, Townsville, Australia
d
School of Nursing, Midwifery and Nutrition, James Cook University, Townsville, Australia
Received 14 June 2012; received in revised form 12 December 2012; accepted 17 December 2012
KEYWORDS
Type 2 diabetes
mellitus;
Taiwan;
Nursing;
Educational
intervention;
Experimental study;
Randomised
controlled trial
Summary
Aims: To investigate whether self care behaviours, medical outcomes and quality of life of Tai-
wanese elderly with Type 2 diabetes mellitus (DM) can be improved by delivery of an educational
health care package.
Background: DM is a major health problem in developed and developing countries, with older
adults constituting about half of the diabetic population. Type 2 DM is the most rapidly increasing
chronic disease in Taiwan.
Methods: During 2005 and 2006, Taiwanese elderly with Type 2 DM (n = 500) were randomly
allocated to either an intervention or control group. Data collection using validated instruments
occurred at baseline and 6 months follow-up. Main outcome measures were blood glucose levels
and diabetic complications.
Results: At baseline, 88.4% participants in the control and 78.8% in the experimental group
had a blood glucose level above normal range (p = 0.076); respective results at 6 months were
92.4% for the control group and 60.4% for the experimental group (p < 0.001). The multivariate
adjusted result showed that the intervention group was 11.1 times less likely to have blood
glucose levels above normal (p = 0.002) at 6 months follow-up compared to the control group.
Occurrence of complications was signicantly fewer in the intervention group at baseline and
at 6 month follow-up compared to the control group (baseline: 42.0% versus 82.1%, p = 0.003;
6 month follow-up: 48.4% versus 87.0%; p = 0.006).

Corresponding author. Tel.: +61 7 4042 1391; fax: +61 7 4042 1590.
E-mail addresses: chaoyh211@yahoo.com.tw (Y.-H. Chao), kim.usher@jcu.edu.au (K. Usher), petra.buttner@jcu.edu.au (P.G. Buettner),
colin.holmes1@jcu.edu.au (C. Holmes).
1322-7696/$ see front matter 2013 Australian College of Nursing Ltd. Published by Elsevier Ltd.
http://dx.doi.org/10.1016/j.colegn.2012.12.006
44 Y.-H. Chao et al.
Conclusion: Although overall occurrence of complications remained unchanged, the educational
health care package specically developed for Taiwanese elderly with Type 2 DM improved blood
glucose levels.
2013 Australian College of Nursing Ltd. Published by Elsevier Ltd.
Introduction
Type 2 diabetes mellitus (DM) is a metabolic disorder char-
acterised by chronic hyperglycemia with disturbances of
the carbohydrate, fat and protein metabolism. It is asso-
ciated with reduced life expectancy, signicant morbidity
due to specic DM related microvascular complications,
increased risk of macrovascular complications (ischaemic
heart disease, stroke and peripheral vascular disease), and
diminished quality of life (WHO, 2006). Poor glycemic con-
trol in persons with Type 2 DM mellitus has other serious
consequences related to cognitive function, psychological
status, for example depression, anxiety and stress, and
the likelihood of increased medical complications such as
retinopathy, renal failure, neuropathy with the risk of ampu-
tation, cardiovascular disease, mortality, and increased
medical costs (Wangberg, 2008).
Background
DM is now recognised as a major health problem in devel-
oped and developing countries with approximately 1 in 20
deaths attributable to the disease (WHO, 2008). Recent
WHO Global Burden of Disease estimates predict that the
worldwide burden in adults to be around 366 million by the
year 2030, with approximately two thirds of those persons
expected to live in developing countries (WHO, 2006). DM is
the most rapidly increasing disease in Taiwan (Chang, 2003)
and it is expected that the number of people with DM in
Taiwan will increase to 300 million by the year 2025 (Chou
et al., 2002). Further, people living in the rural areas of
Taiwan have been found to have a higher likelihood of dia-
betes, which may be accounted for in some way by the
higher numbers of Aboriginal people living in those areas
(Chen et al., 2006). World-wide, older adults constitute
about one-half of the diabetic population. In this age group
DM is a serious disease linked to a higher mortality rate and
shorter life expectancy, mostly due to increased atheroscle-
rotic complications (Wandell and Tovi, 2000). The trend is
similar in Taiwan, especially in rural areas (WHO, 2006).
Many people with DM struggle to adhere to recommended
protocols of self-care and blood glucose management
and thus risk the development of harmful complications
(Diabetes Prevention Programme Research Group, 2002;
Huang et al., 2004). The United Kingdom Prospective
Diabetes Study (UKPDS), a randomised, prospective, mul-
ticentre trial, indicated that improved glucose control in
patients with newly diagnosed Type 2 diabetes mellitus
greatly reduces the risk of clinically evident microvascular
complications (Holman, Sanjoy, bethel, Neil, & Matthews,
2008).
It has been well recognised that in general, between 50%
and 80% of people living with DM have decits in knowledge
and self-care skills (Clement, 1995). As a result, the health
education model (HEM) was used as the theoretical basis for
the study was based on the revised pender health promotion
model (RHPM) (Pender, Caroly, & Murdargh, 2006). Improving
diabetic patient knowledge and changing attitudes to self-
care is one way to work towards achieving better outcomes
for people with diabetes. The model recognises that for
patients to be empowered in self-care, health care providers
must provide the information and skills required for change.
Therefore, a specially designed education programme suited
to the needs of elderly Taiwanese rural dwellers with dia-
betes was developed.
The study
Aim
The aim of the present cluster randomised trial was to inves-
tigate the effects of an educational intervention programme
for Taiwanese elderly with Type 2 DM, many of whom have
low levels of literacy, on blood glucose levels and DM com-
plication rates.
The study addressed two main hypotheses involving
the following two outcome variables: (1) percent of
patients with a well-controlled blood glucose level imply-
ing to have a blood glucose level in normal range (AC:
70110 mg/dl or PC: 90140 mg/dl) and (2) percent of
patients with positive markers for DM complications. Mark-
ers considered were serum creatinine (normal range:
0.61.5 mg/dl), urine analysis (normal: protein and glu-
cose both zero), microalbumin (normal: less than 30 mg),
cholesterol (normal range: 130225 mg/dl), triglyceride
(normal range: 50130 mg/dl), blood pressure (normal
range: 120/80 mmHg, systolic phase is 120 mmHg and dia-
stolic phase 80 mmHg in adult), and cataract or retinopathy
present. Both outcome variables were dichotomised. Ranges
for blood markers were set at current country levels at the
time of the study.
Design
The present cluster randomised controlled trial with follow-
up after 3 and 6 months was conducted in central and
northern Taiwan between 2005 and 2006. The study was
approved by the relevant human ethics committees.
Study protocol
Letters describing the study and seeking permission for
the investigators to phone the potential participants were
mailed from the participating health facilities to persons
who met the sampling inclusion criteria. Each letter con-
tained a self-addressed return envelope allowing interested
persons to send back the signed consent form. Baseline ques-
tionnaires were mailed, and were either returned by mail
after completion (n = 35) or were completed during a face-
to-face interview with the researchers in case participants
Type 2 diabetes mellitusImpact on blood glucose levels and diabetic complications 45
requested help (n = 465). Follow-up questionnaires were
handed out to the participants during their three monthly
medical check-up at the health care facilities. Again, most
questionnaires were completed during a face-to-face inter-
view with the local nurse. There were ten nurses involved
in the study and these nurses were previously trained by
the researchers to ensure consistency in the data collection
process.
Intervention
Both intervention and control participants received a
specially designed information booklet on diabetes. Partic-
ipants in the intervention group were additionally asked
to attend a 1 h diabetes education session every week for
three weeks. In the rst week, patients acquired general
information about DM (introduction to diabetes, signs and
symptoms, treatments, hyperglycemia and hypoglycemia,
and complications). In the second week, patients were
familiarised with specic dietary suggestions (diabetes and
meal planning: a healthy diet, six food groups, how much is
a serving of starch, what are healthy ways to eat starches,
how much to eat each day and how to control body weight).
In the third week patients were educated about the self-care
requirements related to DM (home care of diabetes: exercise
plan, medicine management, self-monitoring of blood glu-
cose, foot and wound care, preparing to travel with diabetes
and introduction to DM support systems in Taiwan). The
information presented was based on available and valid edu-
cational material about DM. The educational material was
delivered in several ways including formal lecturing, role
play, as well as practice and experience sharing to enhance
the learning ability of older people. All educational interven-
tions were delivered within the settings of the participating
health care facilities. The educational interventions were
delivered to six groups of participants ranging between 25
and 70 in size.
Sample/participants
Sample size
Sample sizes of 113 participants were required per group to
detect a difference of 20% between intervention and con-
trol group (adjusted for two tests; Chi-square test; power in
excess of 80%; signicance level 5%). The sample size was
adjusted for the cluster sampling approach (design effect
estimated to be 2) and for losses to follow-up (10%). A sam-
ple size of 250 participants per group was the nal target.
Participating centres
A total of 30 health care facilities (20 randomly selected
local health care centres, ve private clinics and ve
regional hospitals) listed with the non-governmental Dia-
betes Shared Care Networks of Taiwan were approached
for collaboration and 12 agreed (6 local health care cen-
tres, two private clinics and four regional hospitals). These
12 health care facilities became the randomised clusters.
Health care facilities that declined to participate did so
because of administrative requirements which could not be
met by the researcher within the timeframe of the study.
Four physicians, 10 nurses, six staff from the Bureau of Pub-
lic Health of Hsinchu City, and two dieticians collaborated
in the study.
Participants
All patients with Type 2 DM mellitus at the 12 participat-
ing health care services recorded between 2004 and 2005
were considered for participation. Patients who (1) were 50
years or older; (2) were diagnosed as having Type 2 DM melli-
tus (Taiwan ICD-9 code 250) or had poor blood sugar control
(more than 140 mg/dl/AC or more than 200 mg/dl/PC) for
three months; (3) had been treated in the clinic for more
than three months; (4) had functional cognitive-mental sta-
tus and could express personal perception; (5) had no serious
disease or diabetic conditions that may confound diabetic
control or HbA1c value such as acute infections, surgery,
renal disease, liver dysfunction, or low haemoglobin; and
(6) spoke and understood Mandarin or Taiwanese language,
were invited to participate.
Patients diagnosed with a cerebral vascular accident,
end-stage-renal disease, and/or amputations of both legs
were excluded. Originally, 560 patients of the participating
health care facilities were considered eligible. In total 40
participants were excluded, including two persons who did
not meet the inclusion criteria and 38 persons who declined
to participate (Fig. 1). A total of 235 participants in the
intervention group and 250 participants in the control group
completed the follow up questionnaire at 6 months.
Data collection
Validity and reliability
All participants were asked to complete a comprehensive
questionnaire at baseline and at three and six months of
follow-up. The questionnaire included: (1) demographic and
socio-economic data as well as data on the health status
by adapting the Diabetes personal data sheet proposed
by Chang (2003) with permission; (2) The RAND Social
Health Battery a validated questionnaire that measures
the social interaction and social supports of the partici-
pant (Abdulrehman and De Luca, 2001). The scale covers
home and family, friendships, and social and community
life using both open-ended and forcedchoice questions;
(3) Questions on social services were adapted from the
Structured Interview Guide (Ervin, 2004). The ques-
tions focused on the familys experiences of using services
from various agencies and social services; (4) The Health
Perceptions Questionnaire (Ware, 1976) is a validated self-
report instrument that records perceptions of past, present,
and future health, resistance to illness, and attitude towards
sickness. The questionnaire contains 33 items which form six
subscales: current health (nine items), prior health (three
items), health outlook (four items), resistance to illness
(four items), health worry/concern (ve items), sickness ori-
entation (two items), and 6 other uncategorised items; (5)
Summary of diabetes self-care activities (SDSCA) consisting
of 11 items and 14 additional questions. The dimensions of
SDSCA are diet (4 items), exercise (2 items), self-monitoring
of blood glucose (2 items), foot care (2 items) and smok-
ing termination (1 item) (Anderson and Svardsudd, 1995;
Toobert and Glasgow, 1994; Wen, Shepherd, & Parchman,
2004). The validated version of SDSCA used included ques-
tions about self-care recommendations received from health
care providers and medication related questions; and (6)
The World Health Organisation Quality of Life Taiwan Brief
46 Y.-H. Chao et al.
Assessed for eligibility: n = 560 patients
from 12 health care facilities
Cluster randomisation: n = 520 patients
from 12 health care facilities
Excluded: n = 2 not meeting inclusion
criteria; n = 38 refused to participate
Allocated to intervention group: n = 260
from 6 health care facilities
N = 241 received allocated intervention
[19 patients refused to further participate,
saying they were too busy]
Allocated to control group: n = 260 from 6
health care facilities
N = 259 received allocated control
[1 patient had incomplete medical records]
Lost to follow-up (6 months): n = 6
[6 patients refused to further participate,
indicating they were too busy]
Lost to follow-up (6 months): n = 9
[2 patients refused to further participate,
indicating they were too busy; 7 patients
had died]
Analysed: n = 235 Analysed: n = 250
Figure 1 Flow chart of Taiwanese patients with Type 2 diabetes mellitus considered for participation in the educational inter-
vention study.
version (WHOQOL-BREF-TAIWAN) was used in this study to
measure patients subjective perception of quality of life.
The validated questionnaire consists of 26 original items
and two additional items related to Taiwanese culture (Yao,
Chung, Yu, & Wang, 2002). A panel of expert professionals
who specialised in DM examined each item in the survey
independently: physicians, 8 public health leaders, dieti-
tians, and a counsellor (nurse) who had extensive experience
in providing DM education to patients. The panel examined
both the contents and the responses.
Clinical outcome measures
The clinical data collected from the participants medical
records included diabetes-related and nondiabetes-related
co-morbidities, results from ophthalmoscope examina-
tions, blood tests for cholesterol, triglyceride, glycosylated
haemoglobin and nephropathy assessment such as urine
acid, blood urea nitrogen (BUN), and serum creatinine.
Results of urinary tests for micro albumin, urine protein
and urine glucose were noted. Blood glucose levels were
measured either by the researchers or by staff of the partici-
pating health care facilities at baseline and again after three
and six months during routine medical check-up. Elevated
levels of creatinine, urine analysis (protein and glucose
level), micro albumin, cholesterol, triglyceride, and blood
pressure, as well as the occurrence of cataract or retinopa-
thy were recorded as DM complications.
Pilot study
During December 2005 to January 2006 a pilot study in
two small communities in Taiwan, Hsinchu and Taichung,
was conducted. The purpose of the pilot study was to
explore how diabetes researchers delivered diabetes care
programmes in Taiwan and was used to select variables for
the main study and to inform study procedures. In addition,
as the validated questionnaires found in the literature were
not adaptable directly to the sample of the current study,
the scales were modied for use during this phase of the
research. For the pilot study, a convenience sample of 15
patients 50 years or older with Type 2 DM was recruited. Four
individual interviews with professionals (two physicians and
two dietitians) were conducted to explore their views about
the quality of diabetes care in Northern Taiwan. Two patient
focus groups (eight from the experimental and seven from
the control group) were conducted to explore their percep-
tion about diabetes care. The participants of the pilot study
were followed-up for 4 weeks. The pilot study identied
specic problems and suggestions reported by healthcare
professionals and patients. These qualitative ndings high-
lighted the importance of taking into account community
health and social support as well as the structure and pro-
cess of health care. These ndings guided the selection of
variables for the main study.
Ethical considerations
The study was approved by the relevant human ethics com-
mittees and participation was voluntary.
Data analysis
Data were analysed using the Statistical Package for Social
Sciences (SPSS) version 14.0 and STATA for Windows, release
Type 2 diabetes mellitusImpact on blood glucose levels and diabetic complications 47
8. Descriptive statistics were used to describe background
variables and instrument scores. Depending on the distri-
butions, means and standard deviations (SD), or medians
and inter-quartile ranges (IQR) were calculated as meas-
ures of central tendency and dispersion. Standard bivariate
statistical tests (Chi-square, t-test, and non-parametric
Wilcoxon tests) were used to assess relationships between
outcome characteristics and experimental/control status.
The results of those tests were adjusted for the cluster
sampling approach. Multivariate logistic regression analyses
(adjusted for cluster sampling) were conducted to compare:
(1) patients with well-controlled blood glucose levels with
those that had no well-controlled level after six months
of follow-up, and (2) patients with elevated markers for
DM complications with those that had no elevated markers,
allowing adjustment for potential confounding and the base-
line differences between intervention and control group.
The following variables (at baseline) were considered
during multivariate analyses: Group (intervention or con-
trol), sex, age, education, occupation, ethnicity, living alone
or with partner, health insurance coverage, income, history
of DM, DM complications, other chronic diseases, history of
hospitalisation because of DM, history of participation in
education programme about DM, current treatment for DM,
owning a DM patient passport, RAND social health battery,
previous help by health services, previous nancial help,
education services in community, kind of transportation used
by participant, smoking status, total score of health per-
ception questionnaire, and the total score of quality of
life WHO questionnaire. All categorical variables consid-
ered were dummy coded. Results of multivariate logistic
regression analyses were presented as relative risks, 95%-
condence intervals (95%-CI), and p-values. Throughout the
analysis a p-value less than 0.05 was considered statistically
signicant.
Results
Of the 500 participants at baseline, a slight majority was
female (52.2%) and age ranged between 50 and 80 years.
More than half of the participants were 65 years of age or
older (59.2%). The largest ethnic group participating was
Holo (43.6%); other participants included 25.8% Hakkas,
16.8% Chinese Mainlanders, and 13.8% Taiwanese Aborigines.
Many participants (42.4%) had received only elementary
school education, and 23% said that they had not received
any school education.
About 30.6% of participants had been informed of their
diagnosis of DM more than 10 years ago, 27.8% knew for
510 years, 26.2% knew for 25 years, and 15.4% knew
for less than 2 years. The majority of participants (82.2%)
were receiving oral hypoglycemic medication as the sole
treatment. Nearly half of the participants (48.6%) reported
that they previously had experienced at least one com-
plication resulting from their DM, namely cataract 24.2%,
nephropathy 20%, heart disease 17.4%, retinopathy 16.4%,
foot problems 11.4%, and kidney disease 8.4%. More than
half of the participants (61.2%) reported that they also
suffered from other chronic diseases. Seventy-eight par-
ticipants (15.6%) reported that they had been hospitalised
in the previous year for intensive DM treatment or for DM
complications, with most participants reporting a length of
stay between one to two weeks (6.2%).
At baseline, demographic and disease-related character-
istics were generally similar for patients in the intervention
and control groups (Table 1), although patients in the con-
trol group reported no signicant tendencies of longer and
more severe DM in comparison to the intervention group.
Blood glucose levels
At baseline the mean blood glucose level (AC) was signif-
icantly higher in the control group (167.3 versus 147.4;
p = 0.048, t-test), and 88.4% in the control group and 78.8% in
the experimental group had a blood glucose level above nor-
mal range (p = 0.076, Chi-square test) (Table 2). At 6 month
follow-up, the mean level in the control group was 174.4
(50.1) compared to 128.8 (41.0) in the intervention group
(p < 0.001, t-test); while 92.4% in the control group and
60.4% in the experimental group had a blood glucose level
above normal range (p < 0.001, Chi-square test). Multivari-
ate logistic regression analysis adjusted for baseline values
showed that patients in the intervention group were 11.1
times less likely to show blood glucose levels above normal
at 6 months follow-up compared to patients in the con-
trol group (p = 0.002) (Table 3). Patients with higher monthly
income were more likely to have elevated blood glucose lev-
els (p < 0.001), while patients of Aboriginal descent were less
likely to have above normal levels (p = 0.032).
Table 3 is the result of a multivariate logistic regression
analysis comparing participants with elevated blood glucose
level to those who had normal levels at 6 months follow up.
This analysis was required because of differences between
intervention and control group at baseline.
Hence the comparison between the intervention and
control group at follow up had to be adjusted for these
differences. The results showed that despite those differ-
ences at baseline, the intervention reduced the risk of an
elevated blood glucose level at follow-up signicantly in the
comparison to the control group (p = 0.002).
Markers for DM complications
The control group (median number 1; IQR = [1, 2]) had
signicantly more positive markers for complications at
baseline compared to the intervention group (median num-
ber 0; IQR = [0, 1]; p = 0.010, non-parametric Wilcoxon test)
(Table 4). At baseline, 42.0% of the intervention compared
to 82.1% of control group had positive markers for com-
plications (p = 0.003, Chi-square test). The percentage of
patients with elevated triglyceride showed the greatest
difference between intervention and control group (29.0%
versus 60.3%). At six months of follow-up the control group
(median number 2; IQR = [1, 2]) had signicantly more posi-
tive markers for complications than the intervention group
(median number 0; IQR = [0, 1]; p = 0.002, non-parametric
Wilcoxon test). At follow up, 48.4% of the intervention com-
pared to 87.0% of control group had positive markers for
complications (p = 0.006, Chi-square test). No further com-
parisons were made, as the direction of change for the
intervention group was contrary to the initial hypothesis.
48 Y.-H. Chao et al.
Table 1 Comparison of demographic and disease related characteristics between experimental (n = 241) and control group
(n = 259) Taiwanese patients with Type 2 diabetes mellitus at baseline.
Experimental group (n = 241) Control group (n = 259) p-Value
a
Demographic characteristics
Male 48.6% 47.1% p = 0.836
Age 70 years or older 43.6% 38.2% p = 0.175
No school education 20.3% 25.5% p = 0.534
No current occupation 69.3% 68.0% p = 0.556
Married 83.0% 66.8% p = 0.030
Ethnicity Holo 46.1% 41.3% p = 0.902
Language Holo 41.5% 38.6% p = 0.836
Living alone 6.6% 28.6% p = 0.056
National health insurance only 73.0% 73.4% p = 0.893
Median monthly income [IQR]
b
6000 [6000, 18000] 6000 [6000, 15000] p = 0.681
Car available for transport 13.3% 21.2% p = 0.095
Diabetes related characteristics
Diagnosed more than 10 years ago 29.1% 32.1% p = 0.949
Reported diabetes complications 42.3% 54.4% p = 0.326
With other chronic disease 66.0% 57.1% p = 0.366
Not hospitalised for intensive diabetes treatment
during last year
85.9% 83.0% p = 0.643
Participated in prior diabetes education 55.2% 65.3% p = 0.164
Oral medication as sole current treatment 88.8% 76.1% p = 0.090
Diabetes patient passport 71.4% 32.8% p = 0.007
p-value results of Chi-square tests apart from comparison of median monthly income which required a non-parametric Wilcoxon test.
a
Adjusted for cluster sampling strategy.
b
IQR= inter-quartile range.
Table 2 Comparison of clinical characteristics between experimental (n = 241) and control group (n = 259) Taiwanese patients
with Type 2 diabetes mellitus at baseline.
Experimental group (n = 241) Control group (n = 259) p-Value
a
Mean HbA1C (SD)
b
6.6 (1.6) 7.2 (1.7) p = 0.165
Mean Blood glucose, AC (SD) 147.4 (50.3) 167.3 (56.1) p = 0.048
Mean Blood glucose, PC (SD); missing values n = 312 208.7 (76.3) 208.4 (76.9) p = 0.960
With blood glucose level above normal range 78.8% 88.4% p = 0.076
Median serum creatinine [IQR] 0.9 [0.7, 1.1] 1.0 [0.8, 1.3] p = 0.136
Urine protein equal 0; missing values n = 117 100% 80.7% p = 0.246
Urine sugar equal 0; Missing values n = 110 100% 95.4% p = 0.300
Median microalbumin [IQR]
***
; missing values n = 56 0 [0, 0] 0 [0, 1.5] p = 0.030
Mean cholesterol (SD) 190.4 (36.2) 194.8 (46.1) p = 0.570
Mean LDL-C (SD); missing values n = 308 129.6 (33.5) 124.1 (34.9) p = 0.567
Mean HDL-C (SD); missing values n = 291 45.3 (16.1) 50.1 (17.1) p = 0.466
Median triglyceride [IQR] 103 [84, 159] 152.5 [107.8, 193] p = 0.070
Mean systolic blood pressure (SD) 131.0 (15.1) 135.4 (13.7) p = 0.195
Mean diastolic blood pressure (SD) 80.6 (8.5) 80.9 (9.7) p = 0.924
Ophthamalscopic assessment normal 92.4% 77.9% p = 0.021
Mean body weight (SD) 62.7 (11.0) 64.7 (10.0) p = 0.306
With positive marker for diabetic complications 42.0% 82.1% p = 0.003
***
IQR= inter-quartile range, p-value results of Chi-square tests when proportions were compared, t-tests when mean values were
compared and non-parametric Wilcoxon tests when median values were compared.
a
Adjusted for cluster sampling strategy.
b
SD= standard deviation.
Type 2 diabetes mellitusImpact on blood glucose levels and diabetic complications 49
Table 3 Result of multivariate logistic regression analysis of inuence of intervention on blood glucose (normal versus not
normal level; normal level AC 70-110 or PC 90-140) at 6 months of follow-up. Results are based on 485Taiwanese Type 2 diabetes
mellitus patients and were adjusted for the cluster sampling approach.
Blood glucose at 6 months follow-up
Characteristic Normal (n = 112) Above normal (n = 373) Relative risk [95%-CI]
a
p-Value
Control 19 (17.0%) 231 (61.9%) 1
Intervention 93 (83.0%) 142 (38.1%) 0.09 [0.03, 0.33] p = 0.002
Blood glucose level at baseline
Normal 47 (42.0%) 31 (8.3%) 1
Above normal 65 (58.0%) 342 (91.7%) 9.0 [3.8, 21.3] p < 0.001
Average monthly income Continuous 1.01 [1.00, 1.01] p < 0.001
Occupation
None 72 (64.3%) 257 (68.9%) 1
Yes 40 (35.7%) 116 (31.1%) 0.53 [0.25, 1.1] p = 0.089
Aboriginal Taiwanese
No 89 (79.5%) 328 (87.9%) 1
Yes 23 (20.5%) 45 (12.1%) 0.43 [0.21, 0.92] p = 0.032
a
95%-CI = 95%-condence interval; The model was adjusted for the confounding effects of gender (confounded occupation); Occupation
was signicant before adjustment for gender (p= 0.047).
Discussion
The present study was a cluster randomised controlled clin-
ical trial of an educational intervention designed to improve
self-care of older patients with DM in Taiwan. The education
intervention was designed specically to meet the needs
of elderly Taiwanese, including those of Indigenous back-
ground. All materials were presented in an easy to read
format and the sessions delivered to the intervention group
utilised multiple teaching techniques. Riley and McClaughlyn
(2009) pointed out that education is of the essence in pre-
venting DM complications. DM imposes life-long demands on
people with DM and their families, who have to make a mul-
titude of decisions related to managing DM. Patients with
DM need to monitor their blood glucose, take medication,
exercise regularly and adjust their eating habits. Further-
more, they may have to face issues related to living with the
complications of DM. Poor management will result in poor
health outcomes and an increased likelihood of developing
complications. However, most people with DM in Taiwan nd
it difcult to access education due to factors such as cost and
the lack of appropriate services. Previously published inter-
ventions focus primarily on lecture type teaching (Jenum
et al., 2006; Marina et al., 2004; Miller, Edwards, Kissling, &
Sanville, 2002; Mulcahy et al., 2003). The present interven-
tion provided in depth education on all matters related to
DM using a variety of educational approaches. However self-
care and nursing care for DM is a matter that needs to be
incorporated into the routine activities of daily living. The
health care package designed for the study included ses-
sions on how to do self-care and nursing care required for
people with DM, thereby differing from previous diabetes
education programmes in Taiwan. While similar programmes
have been implemented elsewhere and successfully reduced
Table 4 Description of markers for complications for Type 2 diabetes mellitus. Results are based on 500 Taiwanese Type 2
diabetes mellitus patients.
Markers for complications Baseline 6 month follow-up
Control
(n = 259)
Intervention
(n = 241)
Control
(n = 259)
Intervention
(n = 241)
Elevated serum creatinine (normal: 0.61.5 mg/dl 16.4% 4.5% 10.6% 3.1%
Elevated urine analysis (normal: protein and glucose
both zero)
17.1% 0% 23.7% 0%
Elevated microalbumin (normal: < 30 mg) 0% 0% 0% 0%
Elevated cholesterol (normal: 130225 mg/dl) 28.5% 13.5% 17.8% 10.8%
Elevated triglyceride (normal: 50130 mg/dl) 60.3% 29.0% 71.8% 33.7%
Elevated blood pressure (normal: < = 160 to 95 mmHg) 7.3% 5.8% 16.8% 6.0%
Cataract or retinopathy present 22.1% 7.6% 21.0% 7.5%
Patients with at least one marker present 82.1% 42.0% 87.0% 48.4%
50 Y.-H. Chao et al.
blood glucose levels (Pibernik-Okanovic, Prasek, Poljicanin-
Filipovic, & Pavlic-Renar, 2004; Trento et al., 2004), this has
not previously been tried with older Taiwanese. Quantitative
data analysis showed that the delivery of the educational
health care package resulted in improved control of blood
glucose levels. This nding is extremely important as blood
glucose level is one of the main indicators of diabetes
control and is hence strongly implicated in DM complica-
tions (Braiotta, 2007). Strategies that assist with the control
of blood glucose include diet, exercise and taking pre-
scribed medication. In this study the data revealed that the
experimental group were more likely to have followed the
guidelines for a balanced diet, more likely to have partici-
pated in 30 min of exercise during the previous week, and
more likely to have taken their medication as prescribed.
Despite the improvements in blood glucose levels, the
markers for DM complications suggested a similar increase
in the intervention as well as the control group during the 6
months follow-up period, negating the initial hypothesis of
the study. This nding was confounding but may be partially
explained as the rate of complications varied signicantly
between the intervention and control group at baseline. It
would be reasonable to expect that with improved blood
sugar levels, a reduction in the rate of complications would
be evident. This was not the case in this study. However,
it was found that the intervention group developed sig-
nicantly less new complications compared to the control
group.
Several limitations of the current study are worth not-
ing. Firstly, for reasons of practicality, the study applied a
cluster-randomised design. Although the cluster randomi-
sation strategy seemingly worked for most characteristics,
markers for complications of DM were hugely different
between intervention and control group at baseline, which
may have been affected by the sampling strategy. However,
all analysis conducted was adjusted for the cluster sampling
strategy and analyses of the main hypotheses were con-
ducted using multivariate models to allow for adjustment
of confounders. Secondly, due to time and budget consid-
erations, the education sessions were only conducted three
times with each intervention participant. As the participants
were all elderly, different results may have occurred if there
had been the opportunity to conduct additional education
sessions.
Conclusion
The aim of this study was to investigate the effects of an
educational intervention programme for Taiwanese elderly
with Type 2 DM mellitus on blood glucose levels and dia-
betic complication rates. The most striking nding was that
the specially designed programme resulted in a signicant
improvement in blood glucose levels in the intervention
group. These ndings are important because elevated blood
glucose levels are implicated in the development of compli-
cations. Despite this positive nding, there was not a similar
reduction of complication rates in the intervention group.
Further studies are required to support the current nd-
ings, to translate the results to younger patients with DM,
and to test the intervention with people of different ethnic
groups.
Implications for future research
1. The ndings of this study show that a specially designed
health care package is an important method for changing
the level of diabetes knowledge and self care behaviours
for older Taiwanese people with Type 2 diabetes mellitus.
2. Further examination of the relationship between dia-
betes knowledge and self care behaviours is needed over
longer periods of time to determine whether an educa-
tion and skill development programme can help to reduce
complication rates.
Implications for nursing practice
(1) A health care package such as the one used in this study
should be included in the community health promotion
plan when managing a patient with Type 2 diabetes mel-
litus.
(2) An education programme should be offered to the entire
family in countries such as Taiwan where the older com-
munity members tend to live with their family. That way
the family can assist with changes required to enhance
diabetes control.
Conict of interest statement
The authors have no conict of interest to declare.Funding
statement
The study was nancially supported by a PhD scholarship
from James Cook University for Y-HC.
Acknowledgements
The authors would like to thank all Taiwanese participants
and the colleagues of Y-HC in Taiwan. Thanks must also go
to the participating nurses who assisted with the data col-
lection phase of the project.
References
Abdulrehman, R. Y., & De Luca, R. V. (2001). The implications of
childhood sexual abuse on adult social behavior. Journal of Fam-
ily Violence, 16(2), 193203.
Anderson, D. K. G., & Svardsudd, K. (1995). Long-term glycemic
control relates to mortality in Type 2 diabetes. Diabetes Care,
18(12), 15341543.
Braiotta, R. (2007). Diabetes education programmes are essential.
Ala Moana Pacic Center.
Chang, S.-C. (2003). Contributions of patient characteristics and
organizational factors to patient outcomes of diabetes care in
Hualien, Taiwan. Unpublished doctoral dissertation, University
of Texas at Austin: Texas.
Chen, H. F., Ho, C. A., & Li, C. Y. (2006). Age and sex may signif-
icantly interact with diabetes on the risks of lower extremity
amputation and peripheral revascularization procedures: Evi-
dence froma cohort of a half-million diabetic patients. Diabetes
Care, 29, 24092414.
Chou, P., Tung, T. H., Li, C. L., Chuang, S. Y., Li, C. H., Lin, C.
H., et al. (2002). Epidemiology of diabetes mellitus in Taiwan.
Taiwan Journal of Public Health, 21(2), 8396.
Clement, S. (1995). Diabetes self-management education. Diabetes
Care, 18, 12041214.
Type 2 diabetes mellitusImpact on blood glucose levels and diabetic complications 51
Diabetes. (2002). Prevention program research group, reduction
in the incidence of Type 2 diabetes with lifestyle interven-
tion or metformin. New England Journal of Medicine, 346(6),
393403.
Ervin, N. E. (2004). Assessing interagency collaboration through
perceptions of families. Journal of Community Health Nursing,
21(1), 4960.
Holman, R. R., Sanjoy, P. K., bethel, M. A., Neil, H. A., & Matthews,
D. R. (2008). Long-term follow-up after tight control of blood
pressure in Type 2 diabetes. New England Journal of Medicine,
359, 15651576.
Huang, C. L., Wu, S. C., Jeng, C. Y., & Lin, L. C. (2004). The ef-
cacy of a home based nursing program in diabetic control of
elderly people with diabetes mellitus living alone. Public Health
Nursing, 21(1), 49.
Jenum, A. K., Anderssen, S. A., Birkeland, K. I., Holme, I., Graff-
Iversen, S., Lorentzen, C., et al. (2006). Promoting physical
activity in a low-income multiethnic district: Effects of a com-
munity intervention study to reduce risk factors for Type 2
diabetes and cardiovascular disease: A community intervention
reducing inactivity. Diabetes Care, 29, 16051612.
Marina, T., Pietro, P., Enrica, B., Marco, T., Marina, B., Franco, C.,
et al. (2004). A 5-year randomized controlled study of learning,
problem solving ability, and quality of life modications in peo-
ple with Type 2 diabetes managed by group care. Diabetes care,
27(3), 670676.
Miller, C. K., Edwards, L., Kissling, G., & Sanville, L. (2002). Nutri-
tion education improves metabolic outcomes among older adults
with diabetes mellitus: Results from a randomized controlled
trial. Preventive Medicine, 34, 252259.
Mulcahy, K., Maryniuk, M., Peeples, M., Peyrot, M., Tomky, D.,
Weaver, T., et al. (2003). Diabetes self-management education
core outcomes measures: Technical review. Diabetes Educator,
29(5), 768803.
Pender, N. J., Caroly, L., & Murdargh, M. A. P. (2006). Health promo-
tion in nursing practice (5th ed.). Prentice Hall: Upper Saddle
River.
Pibernik-Okanovic, M., Prasek, M., Poljicanin-Filipovic, T., & Pavlic-
Renar, I. (2004). Effects of an empowerment-based psychosocial
intervention on quality of life and metabolic control in type
2 diabetic patients. Patient Education and Counseling, 52(2),
193199.
Riley, P., & McClaughlyn, K. (2009). Diabetes in society: Understand
diabetes and take control: World Diabetes Day 2009. Diabetes
Voice, 54(2), 3839.
Toobert, D. J., & Glasgow, R. E. (1994). Assessing diabetes self-
management: The summary of diabetes self-care activities
questionnaire. In C. Bradley (Ed.), Handbook of psychology and
diabetes: A guide to psychological measurement in diabetes
research and practice (pp. 351375). Philadelphia: Harwood
Academic.
Trento, M., Passera, P., Borgo, E., Tomalino, M., Bajardi, M., Cav-
allo, F., et al. (2004). A 5-year randomized controlled study of
learning, problem solving ability, and quality of life modica-
tions in people with Type 2 diabetes managed by group care.
Diabetes Care, 27(3), 670675.
Wandell, P. E., & Tovi, J. (2000). The quality of life of elderly dia-
betic patients. Journal of Diabetes and Its Complications, 14,
2530.
Wangberg, S. C. (2008). An Internet-based diabetes self-care inter-
vention tailored to self-efcacy. Health Education Research,
23(1), 170179.
Ware, J. E. (1976). Scales for measuring general health perceptions.
Health Services Research, 11(4), 396415.
Wen, L. K., Shepherd, M. D., & Parchman, M. L. (2004). Family
support, diet, and exercise among older Mexican Americans with
Type 2 diabetes. The Diabetes Educator, 30, 980993.
World Health Organization (WHO). (2006). Report of a
WHO/IDF consultation. Denition and diagnosis of dia-
betes mellitus and intermediate hyperglycemia. Retrieved
November 22, 2007, from http://www.who.int/diabetes/
publications/Denition%20and%20diagnosis%
20of%20diabetes new.pdf.
World Health Organization (WHO). (2008). Diabetes Action
Now booklet. A life-threatening condition. Available at
http://www.who.int/diabetes/BOOKLET HTML/en/index3.html
Yao, G., Chung, C. W., Yu, C. F., & Wang, J. D. (2002). Development
and verication of validity and reliability of the WHOQOL-BREF
Taiwan version. Journal of the Formosan Medical Association,
101, 342351.