How public and private health insurance coverage mitigates catastrophic health expenditures in Republic of Korea

Since the early 2000s, it has become a major trend that the health care systems of developed nations move to a mixed version of public and private insurance systems [1‐3]. This is because compulsory social insurance for essential packages of health care services alone cannot satisfy all medical needs, and it is challenging for households to bear the burden of high medical costs for other non-essential health care services. Therefore, many countries with public health care systems have introduced supplementary private insurance, topping up any remaining services with copayments [4].

Republic of Korea (hereafter, Korea) has introduced a national health insurance (NHI) scheme that includes the compulsory coverage of 97% of the population, except those recipients of Medical Aid that protect the accessibility of care for the poor [5]. However, the benefit coverage of NHI is rather low, indicating that the proportion of out-of-pocket (OOP) payments, including copayments for services that have been insured and full payments for uncovered services, is approximately 32.2% of the health expenditure in 2018. This metric is relatively higher than those of Japan (13%), Germany (12.6%), the UK (15.9%), and France (10.2%) [6]. If OOP payments increase excessively, catastrophic consequences for households and the economy may ensue [7]. The World Health Organization (WHO) [8] states that if the ratio of OOP expenses to a household’s ability to pay exceeds a specific threshold, it is considered as “catastrophic health expenditures (CHE),” and this has been adopted as a measure of fairness in financial contribution indicators [8, 9]. Consequently, many studies on CHE have been conducted in Korea for more than a decade, and almost all of these studies have criticized the financial functioning of the Korean NHI scheme, which barely protects households from high OOP expenses [10‐12].

The pricing system of the health care service market in Korea is based on a fee-for-service scheme and NHI is a third-party payer that covers some proportion of medical fees. There are many services in the medical market, most of which are “covered” services managed by NHI, and other “non-covered” services. Notably, NHI covers some proportion of medical costs for services that are covered according to the coinsurance rates, and the rest of the expenses become statutory copayments of patients [13]. At the same time, the patients must make full payments for services such as dental prosthetics, vision correction surgery, manual therapy, and other treatments or medicine based on new health technologies. These uncovered services may have clinical evidence for their treatment effects. However, the NHI does not pay for them due to low economic efficiency or the existence of other alternative medical services.

Moreover, private health insurance (PHI) in Korea covers OOP expenses, the sum of statutory copayments, and costs for uncovered services [10]. Notably, PHI mainly sells two types of insurance plans—fixed benefits and indemnity. The indemnity health insurance plan partially reimburses the patient’s OOP payments. The fixed benefit insurance plan pays the precontracted amount in consideration of medical expenses, loss of income, and other expenses in the case of death, disability, or a few critical illnesses. Some of them are combined with private pension plans. Therefore, the WHO does not regard the fixed benefit plan as a component of the health care system because it coverage is beyond that of the health coverage scheme [14].

Many Koreans additionally purchase PHI plans, a supplementary scheme covering services not covered by NHI. Although some variations exist, depending on research data, it has been reported that approximately 65–80% of households have PHI plans [10, 15, 16]. Moreover, PHI premiums have averaged US$ 184.9 per household with PHI per month, which is 2.1 times higher than NHI contributions (US$ 89.9 per month) [16]. Given this difference in premiums, it would be reasonable for households insured with PHI to be able to significantly reduce their OOP expenses. Furthermore, NHI benefits (benefit-in-cash and benefit-in-kind) are the amounts that NHI pays for medical services according to coinsurance rates; PHI reimbursements are a part of the OOP expenses reimbursed by PHI.

Previous studies have argued that poorly-designed PHI systems increase the prevalence of challenging issues like inequality, insuring only young and healthy people, and causing cost escalation [4, 17]. Such studies have also suggested that well-designed PHI systems can help households avoid the financial shock of large OOP expenditures when accessing health care. However, studies that analyze the extent to which PHI relieves households’ economic burden are rare. Previous studies related to PHI have mainly focused on the effects of PHI subscriptions and the increase in health care use, including adverse selections and moral hazard issues [10, 18‐20] as well as care-seeking behavior [21, 22]. Most studies have shown that PHI increases health care use [23‐25]. However, these studies have limitations because they analyze the effects of PHI subscriptions on health care use without considering the possibility of receiving reimbursements and the attendant amount. Additionally, unlike NHI, the indemnity plans of PHI pay only a portion of statutory copayments and make special reimbursement contracts for some expensive medical services that are uncovered by NHI. Furthermore, PHI’s claim process and exact condition of reimbursements are strict. Therefore, only a small percentage of the insured would receive reimbursements. In order to establish the hypothesis that PHI increases medical service or health care use, the possibility that PHI reimbursements significantly reduce the financial burden on consumers must first be determined. However, in the previous research, there has been no indication of the level of PHI coverage in Korea.

Furthermore, previous studies analyzed only the CHE incidence when estimating the economic burden of households due to medical expenses. However, this method had limitations. First, it was calculated only by using OOP expenses relative to household income; thus, the level of health insurance coverage could not be found. Second, as the unit of CHE incidence is the number of households, it was difficult to adequately measure the economic burden and insurance coverage level. Recently, Jung and Lee [26] developed methods by recalculating the incidence and positive gap of CHE to estimate the effectiveness of insurance in covering CHE. Therefore, this study applied the methods of Jung and Lee [26] and aimed to evaluate the design of the PHI schemes in terms of coverage by estimating how significantly the private indemnity health insurance mitigates CHE in Korea. Through this, we will compare the effects of Korea’s NHI and PHI in reducing CHE and suggest directions for enhancing the coverage and role of PHI.

This study evaluated the coverage of PHI for households by applying a modified CHE calculation method and compared it with the NHI coverage in Korea. A total of 3769 out of 5644 households subscribed to the indemnity plans of PHI, and only 246 households received PHI reimbursements. This revealed that NHI reduced health care inequality by providing more benefits to lower-income households. Conversely, the indemnity products of PHI provided reimbursements more to the higher-income households. This could be interpreted as an indication of the income-regressive aspect of PHI. In particular, the contribution of PHI to CHE reduction was relatively low compared to that of NHI in terms of incidence and positive gap indicators. The number and height of the bar graph in Fig. 1 show that the number of households with beneficiaries and the PHI reimbursements, which represents the effects of reduced CHE, is quite small compared to that of NHI.

The findings of the two-part model with the hierarchical analyses in Table 4 are presented as follows: Model 2 is the case whereby NHI coverage (O2_i − O1_i) is added to Model 1. Moreover, NHI coverage had the most influence among all variables in the incidence and positive gap of CHE. When NHI coverage (O2_i − O1_i) was added to Model 2, it offset the effects of other variables, which were significant in Model 1. Educational level, marital status, and job type were significant among the incidence of CHE in Model 1 but not in Model 2. This indicates that NHI effectively reduces the differences in health care expenses according to socioeconomic status. This interpretation can be validated by acknowledging that Korea operates a fee-for-service system, whereby NHI provides benefits according to the amount of health care used. In Model 2, the odds ratio and coefficient values of income decreased overall compared to those in Model 1 (Table 4). This can be interpreted as the maximum OOP expenses policy, which differentiates the burden of health care expenses according to income level, which has an effect to some extent. However, as the maximum OOP expense policy in Korea is only applied to the health care services covered by the NHI, excluding uncovered services, it seems that the difference in influence based on income level may not be completely offset.

Additionally, in the positive gap analysis, Medical Aid was not significant in Model 1 but decreased significantly in Model 2, and the presence of the four major diseases was significantly higher in Model 1 but not significant in Model 2 (Table 4). Medical Aid recipients in Korea pay only $1 or $2 as OOP expenses; thus, their health care expenses are very low compared to those covered by NHI. Therefore, the results of the positive gap, which appeared significantly negative (−), reflected reality more accurately. Second, Korea is implementing a policy (expansion coverage plan for the four major diseases) to lower the ratio of statutory OOP expenses to the total health care expenses to 5% for four specific diseases (cancers, cerebrovascular diseases, cardiac diseases, and rare diseases) that have high mortality and a high probability of causing high health care expenses [30‐32]. Most Korean studies have concluded that the expansion coverage policy for these four major diseases is ineffective when analyzing CHE. However, we consider these results to be biased because the incidence rates do not change significantly. Studies that have analyzed the effects of policy on the four major diseases using OOP expenses or NHI benefits as a dependent variable tend to report that there is a policy effect [30, 33]; however, studies that adopt CHE incidence as a dependent variable tend to report no effect at all [34, 35]. In this regard, Jung and Lee [26] confirmed that the positive gap can be viewed more accurately than the incidence approach when considering policy effects. Overall, the fact that the four major diseases did not appear to be significant in Model 2 could be understood as the lowering of medical cost burdens by NHI.

The changes between Model 1 to Model 2 were dramatic, but not so in Model 3, which added PHI coverage (O3_i − O1_i). Nonetheless, there are four significant results. First, all the regression coefficients for income in Model 3 were larger than those in Model 2, which indicated increasing inequality. In addition, the odds ratio of the four major diseases decreased from Model 2 to Model 3. This decrease can be interpreted to mean that the PHI coverage effect for these diseases exists because they are the covered under the main plans offered by PHI. In addition, the odds ratio of the number of chronic diseases was higher in Model 3 than in Model 2. This can be interpreted to be based on PHI’s non-acceptance of a high-risk group that may have many chronic diseases. Finally, the most important result was that PHI did not significantly contribute to the reduction of CHE. The reasons for the low coverage of PHI seem to be that indemnity insurance plans reimburse only for a small portion of statutory copayments and some uncovered medical services as well as the strict PHI claim process and detailed conditions of reimbursements.

According to the results of this study, most households were subscribed to PHI and paid premiums that were approximately twice that of NHI; however, the level of PHI coverage was rather low. The purpose of supplementary PHI was not only to meet the diversity of medical service demands but also to supplement the limitations of NHI coverage. Given that PHI in Korea is a part of the wider health insurance system, it cannot avoid the responsibility of protecting households. Therefore, it is necessary to enhance the coverage of indemnity insurance. To do so, the following issues should be addressed.

First, private insurance companies need to disclose transparent data. For example, PHI companies in Korea provide only part of their financial information to several PHI associations, and these PHI associations analyze the financial data and report them on the news. However, there is a lack of trust in such public domain disclosures because there are possibilities that the companies may hide certain information. The data on PHI used in this study were also collected from the patients rather than provided by private insurance companies. In particular, the major PHI companies in Korea announced in 2021 that they would significantly increase insurance premiums due to fiscal deficits. However, it is unknown whether the deficit is due to a large amount of reimbursements being paid to subscribers. In fact, it is possible that insurance plan sales performance is reduced because of the COVID-19 social distancing measures, which limit sales conducted by home salespersons, or that insurance cancellation increases due to the subscribers’ financial deterioration during the COVID-19 pandemic.

Second, it is necessary to establish a unified management system for mutual adjustments between NHI and PHI. For instance, in Ireland, PHI is managed by the Health Insurance Authority under the Ministry of Health and Central Bank of Ireland; hence, it is possible to understand and respond to the insurance market accurately [36]. However, in Korea, PHI is managed by the Ministry of Economy and Finance, and the Ministry of Health and Welfare governs NHI. Under this segmented system, managing the PHI market effectively is demanding. Even if an issue of insurance premium increase emerges, it would be impossible to duly determine whether there has been collusion among the PHI companies.

Third, there is no risk equalization scheme in Korea. It had been argued that the PHI plans made in the early 2000s were designed with generous coverage (combined with pensions), thereby causing deterioration of finance for private insurers over time. Now, albeit those plans have expired, the people who have subscribed since the 2000s have gradually become older and started to use more health care services, hence creating financial deficits for PHI. In Ireland, the Health Insurance Authority operates the “Risk Equalization Fund” [36]. This fund, raised through taxes from all the PHI companies, compensates insurance companies with a higher risk from the elderly. Likewise, a risk equalization scheme could be an option for stabilizing the financial soundness of the PHI companies in Korea. In summary, a new approach is necessary to reestablish the protective role of PHI by simultaneously expanding the benefits and solving funding problems.

To date, the level of practical PHI coverage has been filled with knowledge gaps, and this study provides basic data that aim to fill that gap for the first time. Most studies claim that PHI increases health care use when only considering enrollment status [10, 21‐23, 37]; therefore, the level of reimbursements has not been included in the analysis. As a result of estimating the level of PHI coverage in this study, the coverage was found to be insignificant and the use of health care services was not problematic. In addition, it is difficult to assume if this would affect the NHI fund.

This study presents several limitations. First, although PHI is based on individual subscriptions, CHE is calculated at the household level; therefore, the PHI effect is also calculated at the household level. Second, the level of PHI coverage was somewhat underestimated due to the exclusion of fixed benefit insurance, savings insurance, and other types of plans, which were not included in the WHO’s standards for medical insurance. Third, this study did not conduct a longitudinal analysis because the PHI enrollment rate did not change, and the analysis mainly focused on the comparisons of NHI and PHI coverage. Fourth, this study did not consider the endogeneity issues rooted in sophisticated behavioral health economic theories such as adverse selection, favorable selection, and cream skimming. Because this study focuses on comparing the coverage rates of PHI and NHI, it is unnecessary to ensure homogeneity between groups. If follow-up studies investigate whether NHI pays more than PHI, in cases where patients have the same condition, they must consider the endogeneity issues using statistical techniques such as propensity score matching.

This study reveals that PHI barely protects households from CHE. For private insurance to play the role of medical coverage that subsidizes NHI and reduce the burden of medical expenses of households, it is necessary to improve the benefit coverage of indemnity insurance. The Korean private insurance system needs institutional reforms throughout Korea’s private insurance market structure. In this regard, we propose disclosing transparent data for PHI and implementing a unifying management system and a risk equalization scheme.