Regular HbA1c examination and health outcomes in adult patients with type 2 diabetes in South Korea: A retrospective cohort study

Abstract

Concerns have been raised about the effectiveness of using process-centered indicators to assess the quality of diabetes care in Korea. This study aims to examine the factors influencing the performance of regular HbA1c testing and to explore its association with health outcomes, including hospitalization and mortality. We utilized a retrospective cohort design with a 4-year follow-up period, involving 159,452 adult patients newly diagnosed with type 2 diabetes (E11 in International Classification of Diseases, 10th Edition) in 2011. We established a national population database by merging the Korea National Health Insurance (KNHI) claims database and the KNHI Qualification Database of South Korea. The proportion of diabetic patients who underwent regular HbA1c testing at least once a year in the first 3 years was determined to be 33.8%. In comparison, patients who did not receive regular tests during the same period exhibited significantly increased odds of hospitalization (diabetes/CVD/renal, OR, 1.23, 95% CI, 1.12–1.34; diabetes, OR, 1.36, 95% CI, 1.17–1.57). Additionally, this nonpatient group experienced a higher risk of mortality (OR: 1.56, 95% CI: 1.36–1.80). This study supports the positive impact of regular HbA1c testing on health outcomes for individuals with type 2 diabetes. To increase the current 33% rate of regular HbA1c testing, developing patient-customized management policies is essential. Priority should be given to diabetic patients aged 65 or older, living in rural areas, and those belonging to low-income families (medical aid).

1. Introduction

Diabetes is one of the most common chronic diseases, which is experiencing a rapid global rise. Over the past decade, the number of diabetic patients has more than tripled, escalating from 151 million in 2000 to 463 million in 2019. It is projected to increase further up to 700 million by 2045.^[1] Notably, 3 out of 4 individuals with diabetes mellitus falls within the working-age group, and each year, approximately 4 million individuals between the ages of 20 and 79 succumb to diabetes-related complications, resulting in a mounting social and economic burden.^[1]

In Korea also, the prevalence rate of type 2 diabetes among adults was 10.4% as of 2018, with approximately 190,000 new cases diagnosed annually. The total cost of type 2 diabetes treatment, including medication expenses, has been increasing by about 110 billion won ($84 million) each year.^[2]

Major developed countries have been operating active management programs that focus on improving quality, continuity, and care outcomes based on scientific evidence.^[3–6] The Organization for Economic Cooperation Development (OECD) also has been comparing the quality of diabetes care among member countries using quality indicators.^[7]

Glycated hemoglobin A1c (HbA1c) plays a vital role in diabetes management, and thus, is widely used as a representative indicator for the quality of diabetes care in many countries.^[8,9] Korea has been conducting quality assessment initiatives for diabetes care based on HbA1c test indicators merely since 2011. Despite the regular HbA1c testing rate increasing year after year, the hospitalization rate (per 100,000 people) for diabetes patients has not decreased. This has raised concerns that focusing solely on the process indicators of diabetes management would lead to only increased spending without corresponding improvements in health outcomes.^[2]

This study examined the status of regular HbA1c tests, analyzing the factors influencing regular testing and the relationship between regular HbA1c testing and overall health outcomes for newly diagnosed type 2 diabetes patients in 2011 within the National Health Insurance program. Additionally, it aims to identify patient characteristics associated with poor adherence to regular diabetes management. Through this analysis, we intend to assess the effectiveness of diabetes care quality assessment initiatives in Korea and provide evidence to update the existing policies.

2. Methods

2.1. Study design and data source

This retrospective cohort study examined the relationship between regular HbA1c tests and health outcomes over a 4-year-follow-up period for patients newly diagnosed patients with type 2 diabetes (E11, International Classification of Diseases, 10th Edition (ICD-10)) in 2011.^[10] Patients were investigated for regular HbA1c examinations during the first 3 years of the 4-year follow-up, with hospitalization and mortality assessed in the final year.^[3] Newly diagnosed diabetes was defined as the absence of any prior record of a diabetes diagnosis (ICD-10 code, E10-14) within the preceding 6 years (2005–2010). Based on the International Diabetes Federation’s (IDF) Adult Diabetes Survey of OECD countries, the age of 20 years and older was used as the reference age for study subjects.^[1]

The study data utilized the Korea National Health Insurance (KNHI) Claims Database and the KNHI qualification database covering the period from 2005 to 2015. The data were linked using a separate patient number created by encrypting patients’ personal identification numbers to ensure anonymity. Through the health insurance claims data, we were able to identify the date of diabetes diagnosis, whether HbA1c testing had been performed, the number of outpatient visits, the primary medical institution attended, the Charlson Comorbidity Index (CCI), and hospitalization information. In addition, using the KNHI qualification data, we obtained information on patients’ sex, age, place of residence, income level, and date of death.^[3] This study was approved by the Institutional Review Board of the Korea Institute for Health and Social Affairs (KIHASA) (Sejong, Republic of Korea) (IRB No. 2019-44).

2.2. Study population

In 2011, there were 361,617 newly diagnosed cases of type 2 diabetes in adults aged 20 or older. However, concerns about the accuracy of diagnoses based on health insurance claims data prompted the exclusion of certain patients from the analysis. Specifically, individuals who were diagnosed with type 2 diabetes but used outpatient services fewer than 4 times within the 3 years following their diagnosis (a total of 169,361 individuals) were excluded from the final analysis.^[3,11] In addition, patients (a total of 15,975 individuals) who had been diagnosed with severe illnesses such as ischemic heart disease (IHD), cardiovascular disease (CVD), and cancer before their 2011 type 2 diabetes were also excluded. This exclusion was to ensure the clearer investigation into the impact of regular HbA1c testing on diabetes management for future hospitalization and mortality.^[12] Finally, after excluding patients (a total of 4613 individuals) who had died and patients (a total of 12,216 individuals) who had been hospitalized during the 3-year period in which regular HbA1c testing was evaluated,^[3,12] a final study population of 159,452 patients was selected.

2.3. Study variables

2.3.1. Independent variable: regular HbA1c examination

The study’s independent variable is whether or not regular HbA1c testing was conducted. According to the Korean Diabetes Association’s guidelines, HbA1c testing should be performed at least once per year.^[2] For this study, patients who underwent HbA1c testing at least once a year for 3 consecutive years, in line with this standard, were chosen as regular testers.

2.3.2. Dependent variable: health outcomes (hospitalization and mortality)

The health outcomes analyzed in this study were hospitalization and mortality. Hospitalization was measured as a binary variable, indicating whether a patient was admitted to the hospital with a primary diagnosis code linked to diabetes (ICD-10, E10–E14), cardiovascular disease (CVD, ischemic heart disease (ICD-10, I20–I25), stroke (ICD-10, I60–I64)), or renal disease (ICD-10, N10–N12, N15–N19) during the last year of the 4-year follow-up period.^[3,13] We defined mortality, including all causes of death, based on the information in the KNHI qualification database.^[3,10]

2.3.3. Covariates

Sex (male, female), age (20–44, 45–54, 55–64, 65+), Income level (Medical Aid, Decile 1 and 2, Decile 3 and 4, Decile 5 and 6, Decile 7 and 8, Decile 9 and 10), place of residence (Seoul, metropolitan, city, and county), number of ambulatory care visits for first 3 years (4–12, 13–24, 25–36, 37+), main attending medical institution (tertiary general hospital, general hospital, hospital, clinic, public health center), and Charlson Comorbidity Index (CCI, 0, 1, 2+) were controlled as covariates.^[3]

Patients insured by the National Health Insurance (NHI) have their income levels divided into ten equal parts (deciles), where the highest decile represents the highest income level, except for medical aid beneficiaries. The healthcare institution that patients visit the most frequently for ambulatory care for their diabetes during the first 3 years is considered their main attending medical institution.^[3] The Charlson Comorbidity Index (Charlson index) is calculated based on all diagnostic information for 1 year prior to the initial diagnosis of type 2 diabetes in 2011.^[3]

2.4. Statistical analysis

We conducted a chi-square test to examine the differences in test rates and to understand how patient characteristics are associated with regular HbA1c test rates. Additionally, we performed multiple logistic regression analyses to identify the factors influencing the performance of regular HbA1c testing. Finally, we conducted multiple logistic regression analyses, adjusting for several confounding factors, to determine the relationship between regular HbA1c testing and health outcomes such as hospitalization and mortality. These analyses were conducted separately for diabetes-related hospitalization (diabetes/CVD)/renal disease) and diabetes-caused hospitalization. All statistical analyses were performed using SAS statistical software (version 9.3 for Windows; SAS Institute, Cary, NC).

3. Results

3.1. Differences in regular HbA1c examination according to patient characteristics

Of all the patients involved in the study, the percentage of male patients (54.1%) was higher than that of female patients (45.9%). The largest age group was patients aged 45 and 54, accounting for 29.3% of participants (Table 1). Additionally, 55.4% of the patients resided in city and county areas, and those in the highest income bracket (9th-10th percentile) accounted for 26.8%. Most patients (25.5%) had 13 to 24 outpatient visits over 3 years, and 66.8% preferred clinics as their main attending medical institution for diabetes treatment. Furthermore, 45.8% of the patients had a CCI score of 0.

Table 1.

Differences in regular HbA1c examination according to patient characteristics.

	Diabetic patients		Regular HbA1c		P value
	n	%	n	%
Total	159,452		53,866	33.8
Sex
Male	86,273	54.1	30,633	35.5	<0.001
Female	73,179	45.9	23,233	31.8
Age, yr
20 to 44	29,361	18.4	12,000	40.9	<0.001
45 to 54	46,737	29.3	18,228	39.0
55 to 64	42,558	26.7	14,550	34.2
65+	40,796	25.6	9088	22.3
Residence
Seoul	31,542	19.8	12,187	38.6	<0.001
Metropolitan	39,603	24.8	13,482	34.0
City and county	88,307	55.4	28,197	31.9
Income level
Medical Aid	9864	6.2	2701	27.4	<0.001
Decile 1 and 2	24,024	15.1	7867	32.8
Decile 3 and 4	22,301	14.0	7475	33.5
Decile 5 and 6	26,868	16.8	9199	34.2
Decile 7 and 8	33,593	21.1	11,714	34.9
Decile 9 and 10	42,802	26.8	14,910	34.8
No. ambulatory care visits for 3 yr
4 to 12	52,360	32.8	3736	7.1	<0.001
13 to 24	40,696	25.5	15,092	37.1
25 to 36	34,517	21.7	17,438	50.5
37+	31,879	20.0	17,600	55.2
Main attending medical institution
Tertiary general hospital	12,081	7.6	5210	43.1	<0.001
General hospital	21,072	13.2	8119	38.5
Hospital	14,294	9.0	4254	29.8
Clinical	106,558	66.8	35,077	32.9
Public health center	5447	3.4	1206	22.1
Charlson comorbidity index (CCI)
0	73,014	45.8	27,722	38.0	<0.001
1	40,498	25.4	13,408	33.1
2+	45,940	28.8	12,736	27.7

During the 3-year period, 33.8% of patients received regular HbA1c tests at least once per year (Table 1). The rate of regular testing varied depending on patient characteristics, with males (35.5%) having a higher rate than females (31.8%). The highest testing rate was observed in the 20 to 44 age group at 40.9%, compared to other age groups. In terms of residence, patients living in Seoul had the highest testing rate at 38.6%, and those who primarily used tertiary general hospitals had a higher testing rate (43.1%). The regular testing rate increased with higher income levels and more frequent outpatient visits. Patients with a CCI score of 0 had a higher regular testing rate (38.0%).

3.2. Factors affecting regular HbA1c examination

Table 2 presents the results of the multiple logistic regression analysis identifying the factors influencing regular HbA1c testing. The analysis shows that younger age groups have higher odds of receiving regular HbA1c testing compared to elderly patients aged 65 and above (20–44 years: Odds Ratio [OR] 2.70, 95% confidence interval [CI] 2.59–2.80; 45–54 years: OR 2.25, 95% CI 2.17–2.33; 55–64 years: OR 1.84, 95% CI 1.77–1.90) (Table 2). Patients residing in Seoul (OR: 1.52, 95% CI: 1.47–1.57) and metropolitan areas (OR: 1.03, 95% CI: 1.00–1.06) were more likely to receive regular HbA1c testing than those residing in other regions (city and county).

Table 2.

Odds ratios of regular HbA1c examination by patient characteristics.

	OR	95% CI
Sex
Male	1.01	0.98 to 1.03
Female	1.00
Age, yr
20 to 44	2.70	2.59 to 2.80
45 to 54	2.25	2.17 to 2.33
55 to 64	1.84	1.77 to 1.90
65+	1.00
Residence
Seoul	1.52	1.47 to 1.57
Metropolitan	1.03	1.00 to 1.06
City and county	1.00
Income level
Medical Aid	1.00
Decile 1 and 2	1.11	1.05 to 1.18
Decile 3 and 4	1.10	1.04 to 1.17
Decile 5 and 6	1.16	1.09 to 1.34
Decile 7 and 8	1.26	1.19 to 1.34
Decile 9 and 10	1.41	1.33 to 1.49
No. ambulatory care visits for 3 yr
4 to 12	1.00
13 to 24	9.14	8.77 to 9.52
25 to 36	20.74	19.85 to 21.67
37+	30.72	29.35 to 32.15
Main attending medical institution
Tertiary general hospital	7.04	6.47 to 7.65
General hospital	3.94	3.64 to 4.25
Hospital	1.70	1.57 to 1.84
Clinical	1.53	1.43 to 1.64
Public health center	1.00
Charlson comorbidity index (CCI)
0	1.26	1.22 to 1.30
1	1.15	1.11 to 1.19
2+	1.00

CI = confidence interval, OR = odds ratio.

The odds of receiving regular testing increased with higher income levels and more frequent outpatient visits. Patients who primarily used clinics (OR: 1.53, 95% CI: 1.43–1.64), hospitals (OR: 1.70, 95% CI: 1.57–1.84), general hospitals (OR: 3.94, 95% CI: 3.64–4.25), and tertiary general hospitals (OR: 7.04, 95% CI: 6.47–7.65) had higher odds of receiving regular testing compared to those who mainly used public health centers. Regarding CCI, patients with a score of 0 (OR: 1.26, 95% CI: 1.22–1.30) and those with a score of 1 (OR: 1.15, 95% CI: 1.11–1.19) had higher odds of receiving regular testing than those with a score of 2 or higher.

3.3. Regular HbA1c examination and health outcomes

We conducted a multiple logistic regression analysis to examine the relationship between regular HbA1c testing and health outcomes (Table 3). The analysis was adjusted for sex, age, residence, income, ambulatory care visits, main attending medical institution, and Charlson comorbidity index. The results revealed that patients who did not receive regular HbA1c testing over 3 years had higher odds of hospitalization after 3 years compared to those who did receive testing (diabetes/CVD/renal, OR, 1.23, 95% CI, 1.12–1.34; diabetes, OR, 1.36, 95% CI, 1.17–1.57). Furthermore, the odds of death were 1.56 times higher (95% CI: 1.36–1.80) for patients who did not receive regular testing.

Table 3.

Analysis of the relationship between regular HbA1c examination and health outcomes.

	Regular HbA1c examination
	Yes	No
Patients, n (%)	53,866 (33.8)	105,586 (66.2)
Hospitalization
Diabetes/CVD*/Renal disease
n (%)	906 (1.68)	2221 (2.10)
OR (95% CI)	1.00	1.23 (1.12–1.34)
Diabetes
n (%)	324 (0.60)	664 (0.63)
OR (95% CI)	1.00	1.36 (1.17–1.57)
Death (all-cause)
n (%)	297 (0.55)	1158 (1.10)
OR (95% CI)	1.00	1.56 (1.36–1.80)

Adjusted for sex, age, residence, income, ambulatory care visits, main attending medical institution, and Charlson comorbidity index (CCI).

CI = confidence interval, OR = odds ratio.

^*CVD = cardiovascular disease (including both ischemic heart disease and stroke).

4. Discussion

Well-structured medical services are essential for diabetic patients to maintain healthy lives. Implementing standardized clinical protocols strengthens diabetes management.^[14] Many countries, including the United States, the United Kingdom, and Australia, have introduced active management programs to improve health outcomes for diabetes patients. These programs emphasize continuous patient management by placing primary care at the center and assessing the quality of the care provided.^[3]

Regular testing of Hemoglobin A1c (HbA1c) is one of the most commonly used indicators for evaluating diabetes management. HbA1c testing is widely used in many countries as a key assessment of diabetes management. HbA1c indicates the average blood glucose control over the past 2 to 3 months, encompassing both fasting and postprandial blood glucose levels. It provides information about long-term blood glucose status and can reliably predict the risk of diabetes-related complications. Therefore, regular HbA1c measurement is strongly recommended for diabetic patients to evaluate their blood glucose control. The HbA1c test should be performed appropriately according to the clinical situation and treatment strategy. The American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD) recommend that patients with stable blood glucose levels (HbA1c < 7%) should undergo testing at least once every 6 months. In contrast, patients with unstable or unmet blood glucose targets (HbA1c ≥ 7%) should be tested every 3 months.^[8,9,15]

Since 2011, Korea has evaluated the appropriateness of diabetes management in medical institutions using 7 process indicators. Among these, HbA1c testing is considered appropriate if performed at least once a year. The rate of appropriate HbA1c testing (at least once a year) in diabetes patients has consistently increased since the implementation of appropriateness evaluation, rising from 69.0% in 2011 to 83.1% in 2017.^[2] However, while the appropriate testing rate is monitored yearly, the regular testing rate over the medium to long term is not tracked in relation to patients’ health outcomes. It is essential to manage diabetes adequately from the early stages of diagnosis and maintain continuous management over the long term.

According to this study, 33.8% of newly diagnosed diabetes patients received regular HbA1c testing at least once a year for 3 years. The regular testing rate varied according to patient characteristics; younger age, urban residence, higher income level, frequent outpatient visits, larger primary care facility, and lower disease severity were associated with higher rates of regular HbA1c testing. The regular HbA1c testing rate was notably higher among diabetic patients who resided in densely populated urban areas, faced no economic burden for regular outpatient visits and testing, and utilized medical institutions equipped with major testing equipment and relatively large-scale facilities. These findings suggest that social and economic conditions, as well as geographical accessibility, influenced the regular testing rate. This trend is consistent with previous studies.^[16–18]

HbA1c is a crucial test for managing diabetes. It helps detect prediabetes and take measures to prevent it from progressing into type 2 diabetes. It also aids in diagnosing type 2 diabetes and making appropriate treatment decisions.^[19] The ADA and EASD recommend regular HbA1c testing to maintain HbA1c levels below 7% for effective diabetes management. Studies show that patients who adhere to the guidelines for HbA1c testing frequency are 2.43 times more likely to achieve the target HbA1c level than those who do not.^[9] The HbA1c value is a surrogate marker for glycemic control and a key risk indicator for diabetes-related microvascular and macrovascular complications and mortality. Therefore, regular HbA1c testing is crucial to reduce hospitalizations and mortality risk in patients with diabetes.^[20]

As HbA1c testing is being used as a significant evaluation indicator to assess the adequacy of diabetes management in Korea, the number of diabetes patients undergoing regular HbA1c testing has steadily increased yearly. However, the hospitalization rate (per 100,000 people) for diabetes patients remains very high compared to the OECD average (129.3 people in the OECD average and 245.2 people in Korea in 2017).^[2] This discrepancy has raised concerns that Korea’s approach of focusing on process indicators of diabetes management may lead to increased medical expenses due to more frequent testing, without necessarily improving health outcomes.

This study confirms that regular HbA1c testing can significantly reduce hospitalization and mortality rates among patients with type 2 diabetes. Patients who did not receive regular HbA1c tests for 3 years were found to have a 23% higher risk of being hospitalized for diabetes-related conditions such as diabetes, CVD, and renal disease compared to those who received regular tests. Moreover, they had a 36% higher risk of hospitalization for diabetes alone. The study also revealed that patients who did not receive regular testing had a 56% higher mortality risk. These findings emphasize the importance of regular HbA1c testing in improving the future health outcomes of patients with type 2 diabetes.

This study had a few limitations. Firstly, it did not include all diabetes patients. The analysis focused solely on newly diagnosed type 2 diabetes patients. To examine the impact of appropriate management on health outcomes, it was necessary to align the disease characteristics of the study subjects. Differences in disease characteristics between the management and non-groups could obscure the effects of appropriate management intervention. Since type 1 diabetes has different mechanisms and causes of onset compared to type 2 diabetes, and since disease severity can vary significantly with the duration of illness in existing diabetic patients, only newly type 2 diabetic patients were selected to match the disease characteristics of the study subjects as closely as possible.

Second, due to the limitations of the information available in the claims data, the frequency of HbA1c tests was determined, but changes in HbA1c levels could not be confirmed. This made it impossible to ascertain whether HbA1c was well controlled in patients with type 2 diabetes who received regular HbA1c testing. While Korea recommends receiving HbA1c testing at least once a year, the ADA and EASD recommend testing every 6 months (twice a year). Additional research is needed to determine whether the current testing standards in Korea are appropriate and to identify the most suitable testing standards by examining whether differences in HbA1c levels are based on the interval or frequency of HbA1c testing.

Despite these limitations, this study is significant as it is the first to examine the association between regular HbA1c testing and health outcomes over a 4-year follow-up period among all newly diagnosed type 2 diabetes patients in Korea. Particularly for type 2 diabetes, which requires long-term management, regular testing over the mid-to-long term is necessary rather than short-term testing. This study found that regular testing over 3 years effectively improved health outcomes, highlighting the importance of consistently evaluating process measures, such as HbA1c testing, to enhance diabetes management outcomes in various countries.

In conclusion, this study has confirmed that regular HbA1c testing can significantly improve the health outcomes of patients with diabetes. While many countries are working towards developing diabetes treatment guidelines and management programs, it has been reported that many diabetes patients fail to follow these guidelines properly.^[9] In Korea, 83.1% of diabetes patients received HbA1c testing at least once in 2017. However, only 33.8% of patients received regular testing according to the standard over 3 years. It is crucial to narrow this gap in diabetes patient management in the future. Management policies need to be developed, especially for elderly patients (aged 65 and above), patients in rural areas, and low-income patients (Medical aid) who still have low rates of basic testing.

Author contributions

Conceptualization: Jae-Seok Hong, Hee-Chung Kang.

Data curation: Jae-Seok Hong, Hee-Chung Kang.

Formal analysis: Jae-Seok Hong.

Methodology: Jae-Seok Hong, Hee-Chung Kang.

Writing – original draft: Jae-Seok Hong, Hee-Chung Kang.

Writing – review & editing: Jae-Seok Hong, Hee-Chung Kang.

Project administration: Hee-Chung Kang.

Supervision: Hee-Chung Kang.