The data of this study were retrieved from database: An assessment in Quality of Care among Patients Diagnosed with Type2 Diabetes and Hypertension Visiting Ministry of Public Health and Bangkok Metropolitan Administration Hospital in Thailand (Thailand DM/HT) after the permission of the Medical Research Network of the Consortium of Thai Medical Schools (MedResNet).

The Thailand DM/HT evaluation survey was a nation-wide, cross-sectional study aiming to assess outcomes among patients with diabetes visiting public hospitals of the Ministry of Public Health (MoPH), Thailand and private hospitals and clinics in Bangkok was conducted from 2012 to 2013. The main objective of the Thailand DM/HT evaluation survey was to evaluate the status of care and was supported by the Thai National Health Security Office (NHSO).

The healthcare system in Thailand can be categorized in two types comprising (1) healthcare under the MoPH and (2) private healthcare such as private clinics and hospitals. All Thais have healthcare coverage schemes such as the universal coverage scheme, social insurance scheme and government officer scheme. These healthcare schemes are supported by the NHSO. All of the hospitals under the MoPH at all levels, i.e., community (district), general (provincial) and regional nationwide and some private clinics in Bangkok were invited to participate in the study.

A stratified two-stage cluster sampling method proportional to the size was used to select national and provincial representative samples of patients with diabetes in Thailand. The stratified sample was drawn from a subset of all MoPH hospitals in Thailand. For Bangkok, the targeted institutes included all hospitals and clinics under the NHSO. The first level (province) comprised 77 strata while the second level constituted hospitals within each province. The second level was categorized in 5 strata by size, i.e., regional center hospital (> 500 beds), provincial general (middle) hospital (200–500 beds), first level one (F1) (90–120 beds), first level two (F2) hospital (60 beds) and first level three (F3) hospital (10–30 beds). All university medical centers were excluded from this study.

Inclusion criteria for this study comprised patients with diabetes aged at least 18 years receiving hospital medical treatment in hospital, drawn using the specified sampling method, during the previous 12 months. Any patient who had participated in a clinical trial was excluded. Those patients may have received trial medication or placebo, influencing the outcome of the study.

A total of 833 hospitals under the MoPH were categorized as 33 regional hospitals, 83 general hospitals and 717 first level or community hospitals including the first level one (F1) 73 hospitals, the first level two (F2) 126 hospitals and the first level three (F3) 518 hospitals. All regional and general hospitals were selected, as well as 10% of F1 hospitals, 20% of F3 hospitals and 70% of F3 hospitals. This faction was based on the proportion of patient care provided at the various levels of hospitals. Patients with a diagnosis of diabetes mellitus were randomized and registered at each hospital. A standardized case report form was used to obtain the required information from medical records and was sent to the Thailand DM/HT study of the Medical Research Network of the Consortium of Thai Medical Schools (MedResNet) central data management unit in Nonthaburi. Data were retrieved from patient’s medical records, status of diabetes complications and results of laboratory tests.

Data collected included demographics, weight, height, body mass index (BMI), waist circumference, smoking behavior, systolic blood pressure (SBP), diastolic blood pressure (DBP), cardiovascular complications such as left ventricular hypertrophy (LVH), diabetic complications such as diabetic retinopathy (DR) and diabetic nephropathy (DN), blood chemistry data including fasting plasma glucose (FPG), hemoglobin A1c (HbA1c), hematocrit (Hct), hemoglobin (Hb), serum creatinine (Cr), uric acid, lipid profile including total cholesterol (TC), triglyceride (TG), high density lipoprotein cholesterol (HDL) and low density lipoprotein cholesterol (LDL), available electrocardiogram (ECG) data and results, history of anti-hyperglycemic and antiplatelet drug use and glomerular filtration rate (GFR) calculated using the epidemiology collaboration formula (EPI). Our study enrolled those patients with diabetes who were diagnosed and received ongoing medical care in a hospital. Hospitals in Thailand normally use the standard diagnosis and treatment following Thai clinical practice guidelines for diabetes and diagnosis and classification of diabetes mellitus using Diabetes Care, 2010. Diabetes mellitus was defined as FPG ≥126 mg/dl and confirmed by repeat testing at a second visit. Fasting is defined as no caloric intake for at least 8 h [12]. IHD was defined as myocardial infraction or history of coronary revascularization. Diagnosis of myocardial infraction was performed using standard criteria including stable angina and acute coronary syndrome, categorized as ST-T segment elevation myocardial infraction, nonST-T segment elevation myocardial infraction and unstable angina.

Data were coded and entered in the STATA/MP for Windows, Version 12 (Stata Corp LP, TX). Categorical data were presented as number and percentage. Continuous data were presented as mean and standard deviation (SD). Prevalence was analyzed using descriptive statistics and reported as percentage and 95% confident interval. The chi-square test was used to compare categorical data. Continuous data was compared using the t-test. Continuous data were grouped to analyze associated factor using odds ratio (OR). The magnitude of associations was presented as crude ORs with 95% confidence interval. The multivariate analysis was performed using logistic regression analysis and Forward Stepwise (LR) to adjust confounders. Stepwise p-value for entry was 0.05, and p-value for removal was set at 0.10. The Hosmer-Lemeshow goodness-of-fit of the logistic regression models was performed with p-value = 0.494. The complete case analysis and imputation method were used for any missing data. Missing values were imputed based on the means of the complete case. A p-value less than 0.05 was considered statistically significant.

A total of 25,902 patients with diabetes mellitus were enrolled in this study from 2012 to 2013. Average age was 60.6 ± 10.5 years, 8076 (32.2%) were male and 17,836 (68.8%) were female. The average diabetes duration of patients was 7.1 ± 4.7 years while the average HbA1c level was 8.0 ± 2.2%. Baseline characteristics of this study are shown in Table 1. One third of participants lived in northeastern Thailand. In all, 62.5% subjects visited community or first level (F1, F2 and F3) hospitals. In this study, patients with hypertension totaled 29.1%. Prevalence of IHD among patients with diabetes was 3.54% (95%CI 3.32–3.77), increased with older age and was more common among males. Fig. 1 shows a bar graph representing IHD prevalence for every 10 years of age separated in male, female and overall.

Additional file 1 shows univariate analysis results regarding factors associated with IHD. Significant associated factors included older age, being male, health region, hospital level, insurance scheme, religion, occupation, GFR, HT comorbidity, HT duration, LVH, LAE, AF, DM duration, DN, DR and insulin, sulfonylurea, thiazolidinedione and aspirin therapy. Because a number of patients had missing values for HbA1c, this may have decreased the statistical power of the complete cases analysis. As a result, we applied imputation to handle the missing value and the results of multivariate analysis after imputation showed similarity to the original independent associations with IHD. Multivariate analysis showed the risk factors of IHD included age, being male, being in Health Region 4, hypertensive comorbidity and insulin therapy (Table 2).