Diabetes Research and Clinical Practice - Volume 120 - Supplement 1 - October 2016

213 / 244

disease; cerebrovascular disease; and other causes were 22.0,
24.8, 6.2, 11.2 and 31.3%, respectively. The SMRs according to
the each cause of death were 9.73, 1.76, 2.60, 2.04 and 1.89,
respectively.
Conclusion: Approximately 78.0% of the diabetes-related
deaths would not be ascribed to diabetes in Korea. The diabetic
men have higher risk of dying than women, and diabetic
patients have excess mortality when compared with the
general population. For underlying causes of death not listed
as diabetes, malignant neoplasm was the most common
causes of death in Korea.
PJ-03
Impact of intensified frequent clinic visit on glucose control
Shu-Chuan CHEN 1*, Ming-Hsien CHUNG1. 1Division of
Metabolism, Chang Gung Medical Foundation, Taiwan
This protocol aimed at exploring how HbA1C was improved in
different groups of patients with uncontrolled blood glucose
(A1C < 9.9%, A1C < 10.9%, A1C > 11%) after they received care.
The protocol was implemented between January 1, 2015 and
March 31, 2015 at a metabolism clinic of a teaching hospital in
southern Taiwan where patients with Type 2 diabetes
featuring HbA1c ≧9% attended visits on a monthly basis, had
their dose adjusted, and attended diabetes care programs; the
improvement of their HbA1c after three visits was analyzed to
see if < 9% is achieved. Recruitment was done adopting the
purposive sampling approach and a total of 905 patients
showed improvement. Statistical method: Data were analyzed
with ANOVA and it was found that the group of patients with
A1C > 11% before they received intervention (n = 115) showed
the most reduction by 4.5% ± 1.9%. Significant difference was
observed among the three groups (p < .001). The two groups,
with and without diabetic care, were then analyzed again;
improvements did not reach significance in either group. With
the “doctor-adjusted dose regimen”, the two groups, with and
without consistent doses (addition and no addition of doses),
were analyzed; results showed improvement but no difference
in the reduction of HbA1c. This protocol can help patients re-
examine their uncontrolled blood glucose; they should take on
an approach where they receive treatment provided by the
doctor through monthly visits, have their dose adjusted and
take part in diabetes care programs in order to understand the
right countermeasures they should adopt and to further adjust
their attitude and behavior for the ultimate goal of improving
blood glucose control efficacy.
PJ-04
The analysis of diabetic patients’ blood pressure difference
before and after the doctor visit
Wen-Chin LIN1*. 1Kaohsiung Chang Gung Memorial, Chang Gung
Medical Foundation, Taiwan
Purpose: It was found in the study that when subjects focused
on related clues concerning their physical condition and
disorders, it would also have an effect on the action they
took in response to the disease. Data were collected and
analyzed to understand the difference and reflect actual
quality of care.
Methodology: Data were collected over a period of two weeks,
that is, from September 21, 2015 to October 2, 2015. Five
hundred subjects were enrolled. They followed the clinic
procedure by checking in first at the waiting room to have
their blood pressure taken. After the doctor visit, they went to
another room for health education. Then, a health educator
with a nursing background would take blood pressure from
their right hand for two consecutive times (same as that done
at the waiting room) to get a mean value. Data obtained before
and after the doctor visit were registered and analyzed.
Result: Overall, there was significant difference in terms of
presence of a hypertension history or not among the three age
groups and their SBP values obtained at the hospital and at
home (p < 0.05). The DBP values, on the other hand, showed
significant difference in only the group <50 years old between
that obtained at the hospital and that obtained at home
(p < 0.05). For people with hypertension, those with SBP < 140
and >141 could maintain or improve to be <140, accounting for
60.5% of all; and those with DBP < 90 and DBP > 91 could
maintain or improve to be <90, accounting for 93.5%. For those
without hypertension, SBP was improved more significantly
and the maintenance rate was better (85.5%). In other words,
besides measuring blood pressure correctly, conditions and a
mechanism for measuring blood pressure again should be
established to reflect the actual quality of care. Among people
with hypertension, 43% would monitor their own blood
pressure (40% as is shown in the national survey). Data
obtained from self monitoring of blood pressure help ensure
effective safety required for adjusting medication. Therefore,
encouraging patients to measure blood pressure at home is
equally important in both health education and monitoring
of blood glucose.
PJ-05
The impact of hemoglobin A1c on low-density lipoprotein
cholesterol estimation by different formula in diabetic patients
Ching-Yun HU1, Chia-Lin LEE1,2,7*, Wayne H-H SHEU1,3,4,
I-Te LEE1,3, Yuh-Min SONG 1, Jun-Sing WANG1,3, Chia-Po FU1,6,
Yi-Ting TSAI1, Ang-Tse LEE1 , To-Pang CHEN5, Shih-Yi LIN1,3,8.
1Division of Endocrinology and Metabolism, Department of Internal
Medicine, Taichung Veterans General Hospital, 2Department of Public
Health, College of Public Health, China Medical University, Taichung,
3School of Medicine, National Yang Ming University, 4College of
Medicine, National Defense Medical Center, Taipei, 5Division of
Endocrinology and Metabolism, Department of Internal Medicine,
Show-Chwan Memorial Hospital, Chang-Hwa,6 Graduate Institute of
Biomedical Electronics and Bioinformatics, College of Electrical
Engineering and Computer Science, National Taiwan University,
Taipei, 7Department of Medical Research, Taichung Veterans General
Hospital, 8Center for Geriatrics and Gerontology, Taichung Veterans
General Hospital, Taichung, Taiwan
Background: The Friedewald formula (FF) is the main method
for estimation of low-density lipoprotein cholesterol (LDL-C).
Several modified formulae, including Martin and our recently
developed formulae, had been developed in past years to
improve the accuracy of estimated LDL-C. Because LDL-C
control is an important goal for prevention of cardiovascular
disease in diabetes, we aimed to evaluate the accuracy of
various LDL-C estimation formulae in diabetic patients with
different levels of hemoglobin A1c (HbA1c).
Methods: This is a cross-sectional study enrolled outpatients
diabetes subjects who had full lipid profiles examinations,
including measurements of total cholesterol (TC), triglycerides
(TG), high-density lipoprotein cholesterol (HDL-C), and dir-
ectly-measured LDL-C (dLDL-C) in the same blood sample
between January 2004 and October 2014 at two hospitals in
central Taiwan. Patients with TG level greater than 400 mg/dL
or TC greater than 300 mg/dL were excluded. In FF formula,
LDL-C = TC – HDL-C – (TG/5) (mg/dL); Lee and Hu formula:
LDL-C = 0.75×TC – 25 (mg/dL); Martin formula: LDL-C = TC –
HDL-C– (TG/3.1 to 11.9 according to strata-specific median TG:
very low-density lipoprotein cholesterol ratio). The perform-
ance of each formula was compared in different levels of
HbA1c. The concordance was defined by the proportion of
correct classification in each dLDL-C category (<70, 70–99, 100–
129, 130–159, 160–189, ≥190 mg/dL). Accuracy by different
formulae in different HbA1c levels were compared by Chi-
sqaure test.
Results: A total of 31,814 diabetic subjects were included in
analysis. The overall concordance in each dLDL-C category
according to our Lee and Hu formula were 66.6%, 66.8%, 65.2%,
64.1% for HbA1c<6.5%, 6.5–8%, 8 –9%, ≥9 respectively. (P0.004).
The corresponding concordance for FF were 63.6%, 56.8%,
Poster Presentations / Diabetes Research and Clinical Practice 120S1 (2016) S65–S211 S195