Article Text

Download PDFPDF

Original research
Factors influencing optimal diabetes care and clinical outcomes in Thai patients with type 2 diabetes mellitus: a multilevel modelling analysis
  1. Apinya Surawit1,
  2. Tanyaporn Pongkunakorn1,
  3. Thamonwan Manosan1,
  4. Pichanun Mongkolsucharitkul1,
  5. Parinya Chamnan2,
  6. Krishna Suvarnabhumi3,
  7. Thanapat Puangpet4,
  8. Sophida Suta1,
  9. Sureeporn Pumeiam1,
  10. Bonggochpass Pinsawas1,
  11. Suphawan Ophakas1,
  12. Sananon Pisitpornsuk5,
  13. Chalita Utchin5,
  14. Korapat Mayurasakorn1
  1. 1 Population Health and Nutrition Research Group, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand
  2. 2 Department of Social Medicine, Sunpasithiprasong Hospital, Ubon Ratchathani, Thailand
  3. 3 Department of Family and Preventive Medicine, Faculty of Medicine, Prince of Songkla University, Hat Yai, Songkhla, Thailand
  4. 4 Department of Social Medicine, Samutsakhon Hospital, Samut Sakhon, Thailand
  5. 5 Division of Nursing, Siriraj Primary Care Unit, Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand
  1. Correspondence to Dr Korapat Mayurasakorn; Korapat.may{at}mahidol.ac.th

Abstract

Background Increasing levels of poor glycaemic control among Thai patients with type 2 diabetes mellitus (T2DM) motivated us to compare T2DM care between urban and suburban primary care units (PCUs), to identify gaps in care, and to identify significant factors that may influence strategies to enhance the quality of care and clinical outcomes in this population.

Methods We conducted a cross-sectional study involving 2160 patients with T2DM treated at four Thai PCUs from 2019 to 2021, comprising one urban and three suburban facilities. Using mixed effects logistic regression, we compared care factors between urban and suburban PCUs.

Results Patients attending suburban PCUs were significantly more likely to undergo eye (adjusted OR (AOR): 1.83, 95% CI 1.35 to 1.72), foot (AOR: 1.61, 95% CI 0.65 to 4.59) and HbA1c (AOR: 1.66, 95% CI 1.09 to 2.30) exams and achieved all ABC (HbA1c, blood pressure (BP) and low-density lipoprotein cholesterol (LDL-C)) goals (AOR: 2.23, 95% CI 1.30 to 3.83). Conversely, those at an urban PCU were more likely to undergo albuminuria exams. Variables significantly associated with good glycaemic control included age (AOR: 1.51, 95% CI 1.31 to 1.79), T2DM duration (AOR: 0.59, 95% CI 0.41 to 0.88), FAACE (foot, HbA1c, albuminuria, LDL-C and eye) goals (AOR: 1.23, 95% CI 1.12 to 1.36) and All8Q (AOR: 1.20, 95% CI 1.05 to 1.41). Chronic kidney disease (CKD) was significantly linked with high triglyceride and HbA1c levels (AOR: 5.23, 95% CI 1.21 to 7.61). Elevated HbA1c levels, longer T2DM duration, insulin use, high systolic BP and high lipid profile levels correlated strongly with diabetic retinopathy (DR) and CKD progression.

Conclusion This highlights the necessity for targeted interventions to bridge urban–suburban care gaps, optimise drug prescriptions and implement comprehensive care strategies for improved glycaemic control, DR prevention and CKD progression mitigation among in Thai patients with T2DM. The value of the clinical target aggregate (ABC) and the process of care aggregate (FAACE) was also conclusively demonstrated.

  • quality in health care
  • diabetic retinopathy
  • general diabetes

Data availability statement

No data are available. Raw data used in this study, including de-identified patient metadata and test results, are available upon request. Collaboration is also available.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

STRENGTHS AND LIMITATIONS OF THIS STUDY

  • The study assesses various factors, including clinical targets, processes of care and related factors for achieving glycaemic control, preventing diabetic retinopathy and chronic kidney disease progression.

  • The study employs a robust methodological approach, accounting for the clustering effect of primary care units and using mixed models to analyse clustered and multilevel data.

  • The study might not fully consider changes in treatment practices or shifts in patient characteristics over time, which could influence the observed outcomes.

Introduction

The prevalence of type 2 diabetes mellitus (T2DM) is rapidly increasing worldwide, affecting 536.6 million people (10.5% of the global population) in 2021, and the disease does not discriminate based on socioeconomic or demographic status.1 Global diabetes-related health expenditure was estimated at US$966 billion in 2021, projected to reach US$1045 billion by 2045.1 Alarmingly, almost half of all patients with diabetes (239.7 million, 44.7%) were unaware of their condition, with the highest proportion of undiagnosed cases found in Africa (53.6%), the Western Pacific (52.8%) and Southeast Asia (51.3%).2

Diabetes can lead to various complications and increased risk of cardiovascular diseases.1 Optimal quality of care (QOC) is crucial for minimising complications and improving outcomes in patients with T2DM.3 Despite several studies reporting that improved adherence to diabetes care measures does not necessarily result in better clinical outcomes,4 there has been limited focus on the potential advantages of integrating care processes and clinical targets to form a comprehensive measure of optimal diabetes care. The establishment of a QOC protocol for T2DM is crucial for attaining glycaemic control, mitigating disease progression and preventing complications.5 One study revealed that larger hospitals mostly in urban areas tend to provide higher quality diabetes care, with specialised diabetes clinics (SDCs) outperforming general medical clinics. This could be due to SDCs’ inclusion of specialists, better adherence to guidelines among SDC physicians and improved access to facilities for patients with T2DM.6

In real-world clinical practice, identifying individualised factors influencing glycaemic outcomes is essential for improving diabetes management. Previous studies in Thailand have shown unsatisfactory results in patients with diabetes treated by general practitioners, with a high percentage having a body mass index (BMI) over 23 kg/m², low adherence to home glucose monitoring and poor HbA1c levels.7 Additionally, over 69% of patients with diabetes in Thai primary care units (PCUs) have been found to have poor glycaemic control.7

This study aims to investigate factors significantly associated with achieving glycaemic targets, exploring their relationships with various risk factors. The impact of QOC on chronic kidney disease (CKD) progression is also examined. By identifying care gaps and associated factors, this study seeks to contribute to the development of targeted interventions and strategies for improving QOC and clinical outcomes for Thai patients with T2DM.

Methods

Study design and population

A cross-sectional, hospital-based diabetes registry of 2160 Thai patients with T2DM treated at any of four Thai PCUs was conducted during 2019–2021. The four centres comprised urban-based PCU, and suburban-based PCUs. Siriraj Hospital, which is located in Bangkok, was defined as an urban PCU, while Samut Sakhon Hospital in Samut Sakhon Province, Sunpasitthiprasong Hospital in Ubon Ratchathani Province and Hat Yai Hospital in Songkhla Province were defined as suburban PCUs. These hospitals serve as PCUs in urban or suburban areas serving those living within a 10-km radius of the centre. The two study groups (urban and suburban) have similar socioeconomic status; however, there are differences between groups relative to local culture, customs and lifestyle habits. Each participant was assured by informed written consent before the interview and providing medical information.

Patients satisfying all of the following criteria were eligible for inclusion in this study: guideline-based physician diagnosis of T2DM, age >35 years and having had received ongoing care for at least 12 months during 2018 to 2021. Patients with other types of diabetes, pregnant women and those not willing to provide written consent to participate were excluded. In this study, a proportional-to-size stratified cluster sampling approach was used to collect medical record data from patients in four hospitals. The proportion of Thai patients with T2DM who received at least one HbA1c test during the last 12 months was reported to be approximately 40%.7 Using these data, along with an alpha level of 0.05, a margin of error (d) of 0.03 with design effect of 2, and a 5% increase to accommodate incomplete data, we determined the required sample size for our study to be 2160 subjects. To fulfil this requirement, we recruited 983 patients from the urban PCU and 1177 patients from the suburban PCUs (supplementary sample size estimation and online supplemental figure 1). All PCUs managed patients with T2DM according to standard protocols and diabetic treatment guidelines.8 9

Supplemental material

Patient and public involvement

Patients and/or the public were not involved in the design, or conduct or reporting or dissemination plans of this research.

Data collection and measurement

Details specific to our screening programme were previously reported.7 The standardised questionnaire items included demographic data, self-reported comorbidities, family history, treatment, glycaemic control and chronic diabetic complications (including diabetic retinopathy (DR), diabetic nephropathy (DN) and diabetic neuropathy), which were retrospectively reviewed (see supplementary data measurement). Anthropometric data, biochemistries, behaviours including physical activity, smoking and alcohol drinking habit were recorded.

Outcomes

Diabetes complication assessment

Retinopathy diagnosis was performed by a trained technical specialist and an ophthalmologist using standard techniques and protocols,10 using a non-mydriatic seven-field stereoscopic retinal photography (KOWA Nonmyd 8s Retinal Camera; Kowa, Tokyo, Japan). DR was defined as mild, moderate or severe non-proliferative diabetic retinopathy (NPDR) or as proliferative diabetic retinopathy. The estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration equation.11 CKD prognostic staging was assessed according to Kidney disease improving global outcomes 2022 Clinical Practice Guideline for Diabetes Management in Chronic Kidney Disease.11 We identified two types of kidney injuries including diabetes nephropathy and CKD. Diabetic nephropathy: it typically begins with microalbuminuria, which is an early sign of kidney damage characterised by small amounts of protein (albumin) leaking into the urine due to damaged kidney filters. Over time, it can progress to macroalbuminuria (proteinuria), where larger amounts of protein are excreted in the urine, eventually leading to declining kidney function and end-stage renal disease (ESRD). On the other hand, CKD involves a progressive decline in kidney function over time, often characterised by reduced GFR and elevated levels of serum creatinine and blood urea nitrogen. CKD is staged based on GFR levels and the presence of kidney damage, which can be caused by various factors beyond diabetes.11 The CKD stage here was divided into two groups: a group with normal renal function (eGFR ≥90 mL/min/1.73 m²) and a group with decreased renal function (eGFR <90 mL/min/1.73 m²). Additionally, foot examination and a risk of foot ulcers were evaluated by using monofilament (10 ng).

QOC assessment

QOC consisted of achievement of the clinic targets of care (TOC), and access to the processes of care (POC).7 TOC included the ABC of diabetes (A-glycated haemoglobin (HbA1c), B-blood pressure (BP) and C-low-density lipoprotein cholesterol (LDL-C)). POC included the FAACE indicators of diabetes, where F represents Foot exam, A represents HbA1c exam, A represents Albuminuria examination, C represents LDL-C exam and E represents Eye exam. The study also generated several categories of aggregate QOC measures, including AllABC, AllFAACE, All8Q and Num8Q. Achievement of AllABC indicates that all three treatment goals were achieved; achievement of AllFAACE indicates that all five FAACE exam were conducted. Achievement of All8Q indicates that AllABC and AllFAACE were both fully achieved. The final category, Num8Q, is a count variable that represents the total number of the three possible TOC (A, B, and C goals) plus the total number of the five possible POC (F, A, A, C and E exams) that were achieved (range: 0–8). To determine whether patients achieved the ABC clinical TOC, the study considered three criteria: A—HbA1c <7.0% (53 mol/mmol),12 B—BP <130/80 mm Hg and C—LDL-C <100 mg/dL.13 Patients were considered to satisfactorily meet the examination target (FAACE) if they were examined twice in the least 12 months for HbA1c (yes/no), and once in the previous 12 months for cholesterol (yes/no), albuminuria (yes/no), foot (yes/no) and eyes (yes/no).

Rational drug use

This approach used patient management as the foundation, with first-line drug therapies layered on top according to clinical characteristics, followed by the addition of other drugs proven to protect the kidneys and heart. We retrieved antidiabetic agents, antihypertensive agents, lipid-lowering agents and antiplatelet therapies from the electronic medical records.7

Statistical analysis

All data analyses performed in this study were conducted using STATA V.15.0 (StataCorp LLC), and descriptive statistics were used to summarise patient characteristics. Comparisons of categorical data and continuous data were performed using χ2 test and Student’s t-test, respectively. The data included in this study are multilevel in nature, with some covariates measured at the hospital level, and others measured at the patient level. To account for this multilevel structure, we employed mixed effects logistic modelling, which can identify and remedy any potential hospital clustering effect. Multivariate models were adjusted for age, sex, smoking, alcohol, duration of DM and BMI. The selection of additional risk factors or confounders for inclusion in the multivariable models was guided by a significance level of p <0.25 in the bivariate models. The data analysis used the complete-case approach, where patients with non-missing values for all variables considered in the full multivariable model were included. Missing values were replaced with estimated values based on observed data and statistical models using multiple imputation methods, addressing potential biases. A p value of less than 0.05 was regarded as statistically significant for all tests.

Results

General characteristics of study participants

A total of 2160 patients with T2DM were enrolled in the registry. Of those, 54.5% (n=1177) were treated at suburban PCUs, and 45.5% (n=983) were treated at an urban PCU. The mean age of patients was 63 ±11.3 years and 64.9 ±9.3 years for patients from suburban and urban PCUs, respectively. At baseline, approximately two-thirds of patients in both groups had a BMI reflecting overweight or obese status, which increased considerably in both urban and suburban PCUs (79.2% and 79%, respectively). Urban patients exhibit elevated cardiovascular risk factors: reduced exercise, increased alcohol consumption and higher smoking rates (p <0.001, table 1). Nearly half of them achieved optimal glycaemic control with HbA1c <7% and lipid control with TC <200 mg/dL, TG <150 mg/dL and LDL-C<100 mg/dL as much as normal kidney function with urine albumin-to-creatinine ratio (ACR) <30 mg/g (online supplemental table 1). This was associated with the higher rates of uncontrolled BP and dyslipidaemia in urban-treated patients compared with suburban-treated patients (all p <0.001). Concerning diabetes complications, a significant difference was found for DN exams, DR exams and risk of foot ulcer exams between urban and suburban PCUs (p <0.05) (online supplemental table 2).

Table 1

Demographic and clinical characteristics of patients with T2DM compared between those who received care at an urban PCU and those who received care at a suburban PCU

Rational drug use outcomes

A comparison was made regarding the percentage of rational drug use for ACE inhibitors (ACEI) and angiotensin-receptor blockers (ARB) relative to albuminuria and statin usage. Among the medications belonging to the renin–angiotensin blockage class, losartan and enalapril are the most commonly prescribed drugs for the treatment of albuminuria and hypertension in Thailand. In 33.3% of cases, patients with microalbuminuria received ACEI/ARB, while in 7.7% of cases, those with macroalbuminuria received ACEI/ARB. Optimal ACEI and ARB towards albuminuria stage, 28.37% did not take ACEI/ARB drug and 12.78% received a moderate dose at ACEI20 and ARB50 mg per day. The prevalence of inappropriate ACEI/ARB prescriptions towards albuminuria was 15%. The prevalence of inappropriate and suboptimal dosage of ACEI/ARB prescriptions to treat albuminuria was 14.5% and 20%, respectively (figure 1A and online supplemental table 3). Optimal lipid control at LDL-C <100 mg/dL, 16.82% did not take a statin drug and 19.07% received a moderate dose at 20–35 mg/day. The prevalence of inappropriate statins prescriptions towards LDL-C levels was 25%. The prevalence of inappropriate and suboptimal dosage of statin prescriptions to treat LDL-C levels was 22.6% and 23.0%, respectively (figure 1B and online supplemental table 4). Based on criteria, we found the prevalence of inappropriate metformin prescriptions to be 7%, and 2% of patients were prescribed metformin at doses exceeding the maximum limit of 3400 mg/day. In the CKD stage 3b group, 3% of patients received inappropriate prescriptions (online supplemental table 5).

Figure 1

The comparison of rational drug use of ACE inhibitors (ACEI) and angiotensin-receptor blockers (ARB) towards albuminuria and statins towards low-density lipoprotein cholesterol (LDL-C) levels in participants with type 2 diabetes mellitus (T2DM). (A) The percentage of patients with T2DM receiving ACEI and ARB towards different stages of albuminuria. (B) The percentage of patients with T2DM receiving statin therapy categorised by LDL-C levels.

Targets of care and processes of care

Around 40% of patients in both groups achieved optimal glycaemic control (HbA1c <7%). Suburban patients with T2DM had higher rates of reaching targets for AllFAACE (AOR: 1.62, 95% CI 1.35 to 1.72) and AllABC (AOR: 2.23, 95% CI 1.30 to 3.83) compared with urban patients (table 2). Urban patients underwent albuminuria exams less frequently (AOR: 0.04, 95% CI 0.03 to 0.06). Service location influenced quality-of-care outcomes (POC, TOC and All8Q) as shown in the case-mix adjusted model. Suburban patients with T2DM received more foot exams, HbA1c exams, LDL-C exams, AllFAACE exams and achieved lower BP targets (<140/80 mm Hg) and AllABC rates compared with urban patients. Age was significantly associated with all clinical targets. Long-standing diabetes was negatively associated with HbA1c <7%, AllABC and All8Q, but positively associated with eye exams, foot exams and AllFAACE. Hypertensive patients were more likely to achieve LDL-C exams, HbA1c <7%, and LDL-C <100 mg/dL (table 3).

Table 2

Mixed effect logistic regression for a combination of POC and TOC on difference between urban PCU and suburban PCUs

Table 3

Multivariate multilevel logistic regression analysis to identify factors that independently influence quality of care among Thai patients with T2DM

Factors significantly associated with achievement of glycaemic targets

Multivariate multilevel logistic regression analysis revealed that after adjusting for potential confounders, several factors were identified, including ageing, BMI, longer duration of T2DM, systolic blood pressure (SBP), hypertension, triglyceride (TG), LDL-C, albuminuria levels, drugs (ie, metformin, insulin, ACEI, or ARB), FAACE exams and All8Q (figure 2A and online supplemental table 6). For every 5-year increase in age and every 1-year decrease in duration of diabetes, the odds of achieving good glycaemic control increased by 51% and 41%, respectively. Compared with patients with a normal healthy weight, patients with obesity were 0.68-fold less likely to achieve good glycaemic control. Poorly controlled diabetes was associated with hypertension, higher levels of TG and LDL-C, and the presence of microalbuminuria.

Figure 2

ORs (95% CIs) calculated via mixed effects logistic regression analysis to identify factors that independently predict (A) good glycaemic control, and (B) diabetic retinopathy in participants with type 2 diabetes mellitus (T2DM). An asterisk (*) indicates statistical significance (defined as a p value <0.05). All FAACE, indicates that all FAACE eams or processes of care were performed; All8Q, indicates that all processes of care (FAACE exams) were conducted, and that all targets of care (ABC goals) were achieved; ACEI, ACE inhibitors; ACR, urine albumin-to-creatinine ratio; AOR, adjusted OR; ARB, angiotensin-receptor blockers; BMI, body mass index, HT, hypertension; LDL-C, low-density lipoprotein cholesterol; SBP, systolic blood pressure; TC, total cholesterol; TG, triglyceride.

In addition, patients receiving only metformin were 2.08-fold more likely to achieve good glycaemic control compared with those using a combination of oral antidiabetic agents. Similarly, patients receiving ACEI/ARB drugs had a 1.92-fold higher likelihood of achieving optimal glycaemic control. Achievement of the AllFAACE and All8Q targets was two other factors found to be positively associated with achieving good glycaemic control.

Factors significantly associated with DR

The identified factors included age, obesity, T2DM duration, SBP, ACR, AllFAACE and All8Q. Patients aged ≥65 years (AOR: 2.09, 95%CI 1.03 to 4.25; p =0.042), long-standing diabetes (AOR: 2.19, 95%CI: 1.23 to 3.90; p =0.007) and the presence of macroalbuminuria >300 mg/g (AOR: 2.35, 95% CI 1.28 to 5.32; p =0.021) were found to be independently associated with the development of DR. Additionally, participants with obesity had a 2.06-fold higher likelihood, and those with SBP >140 mm Hg had a 3.14-fold higher likelihood of developing DR complications. Achievement of all FAACE exams reduced the likelihood of DR complications by 0.75 times, and achievement of All8Q reduced the likelihood of DR complications by 0.41 times. The development of DR did not exhibit significant associations with urban and suburban areas (figure 2B and online supplemental table 7).

Impact of QOC on CKD progression

The adjusted ORs for the risk factors independently associated with developing CKD are presented in table 4. Each 5-year increase in the duration of T2DM resulted in a 42% increase in the odds of having CKD. Other risk factors for CKD development and progression included individuals with obesity (1.36-fold higher odds), having a history of gout (3.68-fold higher odds), FBG level >130 mg/dL (1.31-fold higher odds), an HbA1c level>7 (1.36-fold higher odds), an LDL–C level >130 mg/dL (1.37-fold higher odds), microalbuminuria (2.09-fold higher odds) and macroalbuminuria (4.99-fold higher odds), as well as high TG and HbA1c combination group (5.23-fold). Regarding medication, patients with diabetes taking 3–5 drugs and those taking more than 5 drugs had a 2.68-fold and 1.7-fold higher likelihood of having CKD, respectively, compared with patients using fewer than three drugs.

Table 4

Mixed effect logistic regression of risk factors for CKD progression in participants with type 2 diabetes

Discussion

The results of this study provide valuable insights into the characteristics of Thai patients with T2DM receiving care in diverse healthcare settings. The study revealed that achieving optimal glycaemic control remains a significant challenge, with only 40% of patients reaching the recommended target of HbA1c <7%.7 14 Suburban patients with T2DM exhibited higher rates of meeting clinical targets compared with urban patients, possibly influenced by unhealthy urban lifestyles.7 Unhealthy lifestyles and environmental challenges in urban areas, characterised by a higher prevalence of alcohol consumption, smoking and low physical activity, contributed to the difficulties in managing T2DM.15 Urban living complexities led to barriers in self-management and increased rates of hypertension and vascular complications.16 These findings suggest that there may be disparity in the provision of care, patient outcomes, knowledge and effective health management between urban and suburban PCUs.15 16 Understanding the diverse nature of T2DM can enhance the ability to predict clinical outcomes and support precision medicine, ultimately resulting in improved care for individuals with T2DM.17

The study identified several factors significantly associated with better glycaemic control. Younger ages and shorter duration of T2DM were positively correlated with achieving good glycaemic control, suggesting that younger patients and those with a shorter duration of the disease may demonstrate better adherence to treatment and management strategies. Similar subgrouping approaches, Preechasuk et al 18 and other studies,19 20 emphasised the importance of considering variables such as age at diagnosis, BMI, HbA1c, TG and high-density lipoprotein cholesterol (HDL-C) in classifying patients with T2DM. Patients with metabolic syndrome with high TG, low HDL-C, poor BP control and long-standing diabetes were found to have the highest risk of developing retinopathy, albuminuria and neuropathy.21 This could be classified into severe insulin-deficiency diabetes (SIDD) and metabolic syndrome diabetes/mild obesity-related diabetes, while patients who recently diagnosed with T2DM were found to have a mild progression of the disease and complications. Obesity was associated with poorer glycaemic control, and hypertension, higher TG and low-density lipoprotein cholesterol (LDL-C) levels, and the presence of microalbuminuria were linked to suboptimal glycaemic control. Contrasting findings in Japan and Germany, where higher literacy rates and increased physical activity have been associated with better glycaemic control, underscore the importance of balanced weight management, behavioural modification and early screening in specific T2DM clusters treated.22 23 This highlights the importance of balanced weight management and behavioural modification in diabetes care and early screening in specific clusters of T2DM classification.

DR and DN are one of the major microvascular complications of diabetes.24 DR can pose a threat to eyesight, even progressing to blindness.25 The relationship between DR and other factors elucidated independent predictors of DR development, including age, obesity, T2DM duration, SBP, ACR, AllFAACE and All8Q (figure 2B and online supplemental table 7). Consistent with previous studies, our findings identified higher levels of HbA1c, longer duration of diabetes, insulin use and elevated SBP as factors associated with DR.17 Furthermore, during the 4-year screening period in our study, we observed an increase in the severity of NPDR, indicating annual progression in DR severity. Conversely, achievement of all FAACE exams and specific care targets reduced the likelihood of DR complications, underscoring the importance of comprehensive diabetes management.

Our results demonstrate that achieving the HbA1c clinical target (<7%) is protective against the progression of CKD and DN, even after adjusting for other covariates. These findings are consistent with a 2014 meta-analysis26 and two cohort studies27 28 that reported higher HbA1c as a risk factor for decreased eGFR, development of CKD, ESRD and all-cause mortality in T2DM and metabolic syndrome patients.18 Our findings also indicate that coexisting high levels of TG and high levels of HbA1c conspire to cause renal dysfunction in patients with T2DM.29 In Vietnamese T2DM population, HbA1c control was not found to be associated with the development of CKD.30 Our study found some sociodemographic factors to be strongly associated with CKD stage progression, including age, obesity, T2DM duration, and drug utilisation pattern. This result is similar to previous studies that reported advancing age and duration of disease to be significantly associated with CKD in patients with T2DM.31 32 We also found TC, LDL-C and gout to be associated with CKD stage progression. Similarly, previous study reported TC and LDL-C to be associated with CKD.33 Concerning the effect of gout comorbidity on CKD, our findings are similar to those reported from studies in Western population31 and from other Asian populations,34 which reported gout comorbidity to be a strong independent risk factor for CKD stage progression. In addition, similar to a previous study, we found that T2DM treatment with insulin is strongly associated with CKD stage progression, with these patients having over two times higher odds of CKD stage progression.35 Apart from managing blood sugar levels and other risk factors, employing particular medications like metformin or thiazolidinedione (TZD) to decrease insulin resistance, or using drugs that can prevent CKD progression, such as sodium glucose transporter 2 inhibitors, may offer advantages for specific patient subgroups, including those with metabolic syndrome. The analysis of the Diabetes Outcome Progression Trial (ADOPT) database demonstrated a positive effect of TZD therapy on HbA1c levels in patients with SIDD.20

The study also examined prescription patterns for ACE inhibitors (ACEIs), ARBs, statins and metformin. Optimal ACE-I and ARB dosages were not consistently prescribed, with a notable prevalence of inappropriate prescriptions towards albuminuria (figure 1A and online supplemental table 3). Similarly, achieving optimal lipid control at LDL-C <100 mg/dL was suboptimal, with a significant proportion of participants not receiving statin therapy or receiving inappropriate doses (figure 1B and online supplemental table 4). Moreover, inappropriate metformin prescriptions were observed. These findings emphasise the importance of implementing guidelines and regular monitoring to ensure appropriate and optimal drug use in the management of T2DM.

This study possesses several notable strengths. The data were collected from a large cohort of patients with T2DM receiving care in urban and suburban PCUs across multiple provinces in Thailand. The comprehensive data collection, large sample size and multilevel analysis enhance the robustness of the study’s findings and provide valuable insights for informing healthcare practices and interventions aimed at improving diabetes care and patient outcomes. However, it is important to acknowledge certain limitations of the study. As a cross-sectional study, causality cannot be established, and the findings provide associations rather than definitive conclusions. The study focused on a specific population in Thailand, which may limit generalisability to other populations. Additionally, the multivariate modelling relied on complete-case analysis, potentially introducing information bias in cases of missing data. Future prospective studies and interventions are warranted to further explore the identified factors and validate the effectiveness of targeted interventions in improving diabetes care and patient outcomes.

Conclusion

The findings of this study highlight the need for targeted interventions to address disparities in care between urban and suburban areas, optimise drug prescription practices and implement comprehensive multifactorial care strategies to improve glycaemic control, prevent DR complications and mitigate CKD progression in Thai patients with T2DM. Moreover, the confirmed value of achieving the clinical target aggregate (AllABC) and the process of care aggregate (AllFAACE) strongly highlights the need to improve comprehensive patient QOC. These findings may also help to influence changes in healthcare policy that expands coverage for important modalities to help achieve care and process targets and improve the overall management of T2DM in Thailand.

Data availability statement

No data are available. Raw data used in this study, including de-identified patient metadata and test results, are available upon request. Collaboration is also available.

Ethics statements

Patient consent for publication

Ethics approval

The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board of Faculty of Medicine Siriraj Hospital, Mahidol University (COA no. Si 330/2017)), Ethics Committee of Sunpasitthiprasong Hospital (076/2562), and the Human Research Ethic Committee of Faculty of medicine, Prince of Songkla University (REC.61-210-9-1). Participants gave informed consent to participate in the study before taking part.

Acknowledgments

The authors gratefully acknowledge the study participants enrolled from and the staff that work at each medical centre included in this study. The authors also graciously acknowledge the following groups for their assistance with data collection: the Department of Social Medicine, Samut Sakhon Hospital (Samut Sakhon Province, Thailand), the Cardio-Metabolic Research Group, Sunpasitthiprasong Hospital (Ubon Ratchathani Province, Thailand), and the Division of Family Medicine of the Department of Community Medicine, Faculty of Medicine, Prince of Songkla University (Songkhla Province, Thailand).

References

Footnotes

  • Contributors AS, TP and KM conceived and designed the study protocols. AP, TP, TM, PM, SS, SP, BP, SO, SP (Sananon Pisitpornsuk), CU and KM performed patient care and data collection. PC, KS, TP (Thanapat Puangpet) and KM supervised the research study. AS and KM wrote both study protocols. AS wrote the data management and statistical analysis sections of the protocol and manuscript.

    AS and KM drafted and finalized the manuscript. All authors reviewed and revised the manuscript for important intellectual content. AS and KM oversaw study implementation, assisted in writing and editing the paper and prepared the manuscript for publication. All authors have read the manuscript and are in agreement with the decision to submit the manuscript for journal publication.

  • Funding This study was supported by grants from Mahidol University (grant no. 10492), and from the Faculty of Medicine Siriraj Hospital, Mahidol University, Bangkok, Thailand (grant no. R016210002).

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.