Abstract
Objectives:
Although Korean Medicine (KM) is an important part of the Korean healthcare system and plays a significant role in chronic and incurable diseases, there is insufficient information about KM utilization. The aim of the study was to identify KM utilization patterns using a national representative sample data.
Methods:
KM users aged 0–80 years and older from the nationally representative sample in the National Health Insurance Service–National Sample Cohort of 778,506 KM users between 2002 and 2013 (were analyzed. Annual changes in utilization patterns were analyzed by sex, nine age groups, five income levels, and numbers of visits and hospitalizations. Utilization patterns by the type of healthcare institution and the top 10 major diseases were studied.
Results:
Frequencies of KM use differed according to analyses of patients and claims. Women used KM 1.5 times more than did men. Patients in their 40s and 50s made up one third of KM users. In contrast to other studies, high-income groups used KM more than did low-income groups. More than 96% of ambulatory patients used mostly KM clinics, and more than 76% of inpatients used KM hospitals. Musculoskeletal disorders were the main disease treated, which conformed to the results of previous studies.
Conclusion:
The results suggest that women, people in their 40s and 50s, and people with a relatively high income use KM more often than other patients. Further comparison studies of both Western medicine in Korea and other countries should be conducted.
Introduction
T
Korean medicine (KM) is a form of traditional medicine developed on its own whole person–centered medicine system. Although KM shares its origins with Chinese and Japanese medicine, KM has unique characteristics and modalities, such as Sasang constitutional medicine, Saam acupuncture, and Chuna. 5,6 The services of KM include acupuncture, moxibustion, cupping, pharmacupuncture, and herbal medicine. According to a study in 2014, 27.9% of Koreans use KM. 7,8 Korea operates a healthcare system that includes both conventional Western medicine and KM. 9 Korea started a mandatory social insurance system in 1977 and extended coverage to the entire nation in 1989. KM has been covered by the National Health Insurance (NHI) system since 1987. 10 More than 97% of the population is covered by the Korea National Health Insurance Program (KNHIP), and the remaining 3% are covered by a medical aid program that is operated by the Korea National Health Insurance Service (KNHIS). 11 However, KM expenses accounted for only 3.9% of KNHIP costs and 2.3% of medical aid program costs according to the total NHI costs for 2013 12 because the KM services covered by NHI are restricted to diagnosis, 56 herbal medicinal extracts, 13 acupuncture, and cupping. Most medication services (herbal medicine) are noncovered items.
Since 1987, the medical services of KM have been expanded quantitatively as well as qualitatively in the national system. Although previous studies have investigated the utilization patterns of KM, including practice patterns, methods of payment, and costs for the entire population, few studies have analyzed KM utilization patterns using national health insurance data. 14 The third Korean medical use and consumption survey, conducted in 2015, did not fully explain KM users' characteristics because of a lack of representativeness of survey data, such as potential recall bias.
This study used a national sample cohort constructed from the National Health Information Database in the NHI system this database was recently constructed to meet an increasing request for public use and the application of the health insurance data recently. 15
Materials and Methods
Data sources and data extraction
The KNHIS covers the entire population with the KNHIP and medical aid program; thus, the KNHIS includes all health insurance records in Korea. 10 The National Health Insurance Service–National Sample Cohort 2002–2013 (NHIS–NSC 2002–2013) is a dynamic cohort constructed in 2013 from the KNHIP. Because it includes long-term observation data, NHIS–NSC can evaluate temporal relationships and causality.
A representative sample of approximately 1,000,000 people was extracted from the 2002 database using a stratified, simple random sampling method. This sampled database consisted of 1476 subsets stratified by sex (2 groups), age group (19 age groups), income level by subscriber type (industrial workers and self-employed; 20 groups), and medical aid (0 groups). Finally, 1,025,340 registrants were included in the database, which represents 2% of health insurance beneficiaries. A total of 778,506 patients (370,435 males and 408,071 females) who visited KM clinics or hospitals at least once were analyzed. All patients' personal information was anonymized.
Study variables
In the study, a KM user was defined as a patient who had at least one KM clinic or hospital visit. Patient ages were categorized into nine groups as follows: 0–9 years, 10–19 years, 20–29 years, 30–39 years, 40–49 years, 50–59 years, 60–69 years, 70–79 years, and 80 years or older. Regarding socioeconomic status, patients were classified with a well-defined monthly income into six categories according to income percentage as follows: the lowest income group (lowest 20%), the low-middle income group (lower 20%–40%-), the middle income group (lower 40%–60%), the middle-high income group (upper 20%–40%), and the high income group (upper 20%).
Number of visit days and cost
Times of visit (outpatient) or days of hospitalization (inpatient), identified from diagnoses classified according to the disease classification code on each patient's medical care fee bill, were used. The total medical expense was the total medical care cost, and the copayment was a beneficiary burdened cost.
Types of healthcare institutions
In the Republic of Korea, total medical expense and beneficiary burdened cost are determined by the types of medical institutions according to size and are categorized as follows: A tertiary general hospital is a training hospital that covers severe cases. A general hospital is a large medical institution with over 100 admissions. Hospitals are defined as hospitals, dental hospitals, KM hospitals, and medical care hospitals that have 30 or more beds. A KM clinic is mainly for outpatients, and a health center is used for community health.
Classification of disease category
The Korean Classification of Diseases (KCD)-6, based on the International Classification of Diseases 10th Clinical Modification, was used to determine disease category. The main diagnosis was used for the classification. The top 10 primary diseases were selected to demonstrate KM claim patterns.
Statistical analysis
The units of observation were each individual or case included in the study sample. Descriptive statistics were used to describe the basic characteristics of the study sample. SAS software, version 9.2(SAS Institute Inc., Cary, NC), was used for data management and analyses.
Results
A total of 778,506 (370,435 males and 408,071 females) KM users between 2002 and 2013 were analyzed.
Characteristics of KM users by year (inpatients and outpatients)
Supplementary Tables S1 and S2 shows annual changes in sex, age, income level, number of visit days, total medical expenses, and beneficiary burdened costs according to inpatients and outpatients. Patients were defined as the number of diseased individuals among beneficiaries, and this value was counted according to disease, month, and medical care institution for the current year.
The number of KM users has increased steadily each year. Women use KM more often than men do. Across all age groups, patients in their 40s used KM most frequently until 2010. After 2011, individuals in their 50s used KM most often because patients in their late 40s aged into their 50s during the study period. Patients in the middle-high income group were the most frequent users of KM.
The mean number of outpatient visit days per person increased from 6 days in 2004–2006 to 7 days in 2005 and 8 days in 2011. The median number of outpatient visit days remained consistent at 3 days. The mean number of inpatient days of hospitalization per capita increased by 32 days in 2008 but decreased to 22 days after 2009. Regarding ambulatory patients, the total medical expense per person was $70 (U.S. dollars) in 2002 and rose consistently to $145 by 2013. By contrast, the mean total amount per capita of inpatients varied before 2008 (mean $574 in 2002 to $1301 in 2008).
Characteristics of KM users by the type of medical care institution
Figure 1 presents the characteristics of KM users by the type of medical care institution. Ambulatory patients used KM clinics most frequently across all years (96% or greater), and at least 76% of inpatients visited a KM hospital. Since 2010, the number of KM visits has been reported as the cross-employment of Western medicine and KM doctors where available.

Characteristics of Korean medicine (KM) users by the type of medical care institution.
Total medical expenses and mean numbers of visits of KM users by income level (outpatient)
Table 1 presents the total medical expenses and average numbers of visits of KM users by income level among outpatients. Since 2002, patients with a higher income level and more insurance have used KM often. Middle-high–income and high-income patients used KM more often than patients in the low-income and medical aid groups. Generally, patients with higher incomes had consistently higher total medical expenses per person. Regarding mean total medical expenses, high-income patients made up the group with the highest expenses at $75 in 2002 and $129 by 2010. Since 2011, the group with the highest mean total medical expenses has been the medical aid group, followed by the high-income group. In addition, the group with the highest mean number of visits from 2004 (6.8) to 2010 (8.3) was the high-income group; however, since 2010, individuals in the medical aid group had visited medical care institutions more frequently, followed by the high-income group.
Medical expenses are expressed as U.S. dollars.
KM, Korean medicine; min, minimum; max, maximum.
Top 10 major diseases for KM insurance claims among inpatients and outpatients
The KCD-Oriental Medicine (KCDOM)-2 was used for the data from 2002 to 2009, and KCDOM-3 was applied to the data after 2010. KCDOM-3 was developed for double coding with Western disease names (International Classification of Diseases). 5,16 The most frequently claimed diseases for outpatients were musculoskeletal diseases, such as backache and sprains, according to KCDOM-2. Stroke and nerve disorders, such as sequelae of stroke, paresthesia, and facial nerve palsy, ranked high, in addition to backache. Since 2010, musculoskeletal diseases, such as dorsalgia and soft tissue disorders, have been claimed most frequently. Supplementary Table S3 shows the top 10 diseases for KM insurance claims for inpatients. The most frequently claimed disease among inpatients each year from 2010 to 2012 was dorsalgia, followed by lumbar spine and pelvic lesions, other intervertebral disc disorders and cerebral infarction. In ambulatory patients, according to Supplementary Table S4, dorsalgia was also claimed most frequently, and soft tissue disorders, shoulder lesions, lumbar spine, and pelvic disorders (sprain and strain, dislocation) and gonarthrosis were also ranked high.
The median number of visit days among outpatients remained consistent at 3 days. The mean number of hospitalization days per capita in inpatients increased by 32 days in 2008 and has decreased to 22 days since 2009.
Discussion
This study identified the utilization patterns of KM by using a national representative sample cohort data and analyzed KM users by sex, age group, income level, and numbers of visit days and hospitalizations. These results indicate that the numbers of KM users and claims increased each year in both inpatients and outpatients. Women used KM more often than men. Patients in their 40s and 50s represented the largest proportion of KM users (approximately 20%). Middle-high–income and high-income patients used KM more than low-income and medical aid patients. The mean numbers of visits were 4.8 in the medical aid group and 7.9 in the high-income group, which implied that the mean of the high-income group was approximately twice that of the medical aid group in 2008.
These results support the report of the Third Korean Medical Use and Consumption Survey in 2015, in which higher-income groups demonstrated more frequent KM use. 7 By contrast, the current study results conflict with Yoon and colleagues' study, which found that the relatively low-income group used KM more frequently compared with other groups. 17 This difference was attributed to data collection methods and the personal income of interviewees included in the survey. In Taiwan, meanwhile, women were the main users of Chinese medicine. Individuals in their 20s and 30s and the low-income group took advantage of Chinese medicine. 18 The income data of the NHIS-NSC were categorized by family income. The current results indicate that the group with the highest total medical expenditure was the medical aid group followed by the high-income group, from 2011 to 2013. This finding is not consistent with previous years. Thus our results do not warrant that the high income group spends more on medical expenses. Therefore, further investigation of the characteristics of KM users by income level is required.
Currently, total medical expenses and beneficiary burdened costs are determined by the types of medical institutions according to size in the Republic of Korea; this study analyzed the characteristics of KM users according to medical care institutions. Patients making KM clinic visits made up at least 96% of outpatient KM users, and patients with KM hospitalizations made up 76.1%–98.6% of hospitalized patients (Fig. 1). The proportion of medical care hospital users decreased consistently between 2008 and 2013 among inpatients; by contrast, the proportion of KM hospital and KM clinic users increased. Among outpatients, the proportion of KM hospital users increased slightly from 2008 to 2013, but this change does not appear to have been significant. The most frequently used institutions were KM clinics, as at least 96% of users visited KM clinics in 2012. According to the NHIS Statistics 2013, the number of KM hospitals gradually increased (146 in 2008 to 212 in 2013, a 45.2% increase). However, the number of KM clinics increased from 11,334 in 2008 to 13,100 in 2013 (15.6% increase), which was a lower rate of increase than that of the KM hospitals. 19 Therefore, the current results appear to have been affected by this trend regarding the number of medical institution. Although KM departments have been established in Western medicine hospitals, these departments have not been used as frequently in Western medicine hospitals as in KM medical institutions.
These patterns of KM users by categorized medical care institutions were somewhat similar to patterns in Taiwan. According to Chen and colleagues' report, most TCM visits were to private TCM clinics (82.6%), followed by private TCM hospitals (12.0%), in outpatients. Western medicine hospitals with TCM departments accounted for only 4.6% of visits. 20 Another study showed that most of the active TCM physicians (82%) worked in personal practice clinics and that only a small portion (18%) worked in TCM hospitals, which supported Chen and colleagues' report. 21
Additionally, the patterns of users of dentistry and Western medicine according to medical care institutions in the Republic of Korea were studied using the NHI Statistics 2013. 17 Of the outpatients, 77% visited clinics for Western medicine, which was a much lower figure than the number KM users who visited outpatient clinics. The patterns of users receiving dental services were similar to those of KM users, as 95% visited dental clinics. Most Western medicine visits were to hospitals (38%), similar to the scale of KM hospitals followed by general hospitals (31%) in inpatients; hospitalizations for dental treatment made up over 99% of dental hospital visits.
Regarding the utilization of KM according to disease, musculoskeletal system disorders represented the most frequently claimed disease in both inpatients and outpatients, which corresponds to the results of the study by Lim and Yoon, which evaluated the NHI data in 2012 and the Report on the Usage and Consumption of Korean Medicine 2011. 17 –22 Compared with the Taiwan study, diseases of the respiratory system, diseases of the musculoskeletal system, and connective tissue injury and poisoning accounted for 22.1%, 18.1%, and 16.1% of claims, respectively. 18,23
The major strength of the current study was that it investigated changes in usage patterns according to income level. The study used long-term cohort data that can be used to predict KM use patterns and user characteristics according to sociodemographic factors. Additionally, this study included a relatively large sample size of KM users.
The limitations of the study were as follows. Because most KM treatments, including herbal decoction prescriptions, are not covered by the KNHIP, determining total KM utilization patterns is difficult. Although the proportion of KM hospitals in total KM utilization was slightly low, when considering the rate of health insurance coverage at a KM hospital (30.8%), the noncovered services at KM hospitals (56.0%) had a slightly higher proportion compared with the noncovered services of Western medicine according to the National Research on Medical Payment in the Republic of Korea 2013. 22 Oh et al. investigated the relative preference for traditional KM services according to disease using data from the 2008 Korea Health Panel. This study reported the numbers of outpatient visits due to patients' diseases and the costs paid directly by patients, including copayments and the cost of care, that are not covered by national health insurance, including the cost of herbal medicines at KM hospitals and clinics. 24 Although this study focused on information about claims did not include noncovered services, which could not be measured accurately compared with Oh and colleagues' previous study, the current study results supported Oh and colleagues' study in terms of preferences for KM for specific diseases. KM use was higher in patients with musculoskeletal system diseases or cerebrovascular disease, which was similar to the results of the study of the Korea Health panel. Therefore, it seems that the limitations of the study were not highly influential.
Another limitation is that the diagnosis code designed for insurance claims may be inaccurate. To estimate the health equity of KM in the insurance system, continuous monitoring and assessments are necessary.
On the basis of this study, further research should be conducted to identify the factors affecting KM use and health status. Specifically, detailed comparison studies with other countries that have a single-payer insurance system, such as Taiwan, are required. 18 –26
Conclusion
This investigation was the first study of KM utilization patterns with a large sample of national representative data from the NHIS-NSC 2002–2013. Although controversy exists regarding utilization patterns according to income level, this study has confirmed the results of previous studies of KM utilization for sex and disease tendency. It seems that these patterns of utilization exist because of the high proportion of KM services that are not included in the NHIS system, which indicates that the KM services covered by NHI are restricted and have low economic accessibility. The results of this study contribute to understanding of KM users and can be used as a basic foundation for related policy decisions made by government officials and medical professionals. Further studies should be conducted to explore the predictors of KM utilization, and comparison studies of Western medicine in the Republic of Korea and other countries should be performed.
Footnotes
Acknowledgments
This work was supported by a grant from Kyung Hee University in 2012 (KHU-20121688).
Author Disclosure Statement
No competing financial interests exist.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
