1. Introduction
In response to the call of the World Health Organization (WHO) and to promote population health status, many countries are exerting effort to achieve universal health coverage (UHC), which is a major Sustainable Development Goal. Achieving UHC means that all people receive essential health services without being exposed to financial hardship [
1]. Since 2003, China has introduced and implemented a comprehensive package of basic medical insurance (BMI). The BMI system consists of urban employee basic medical insurance, urban resident medical insurance (URMI), new rural cooperative medical scheme (NRCMS), and the urban and rural medical assistance system. These basic medical insurance plans cover about 97% of the population (over 1.3 billion Chinese), which enabled China to achieve near UHC [
2].
Despite nationwide coverage, the service provisions and financial protection of the Chinese BMI system are far from sufficient. The risk of various critical diseases (e.g., malignant tumor, acute myocardial infarction, coronary artery diseases, cerebral stroke, etc.) and their economic burden are dramatically increasing with an aging society. Critical illness caused 36.19 million deaths and 662.71 million years of life lost worldwide in 2016 [
3]. The Global Burden of Disease Study 2017 also shows that non-communicable diseases accounted for 73.4% of deaths [
4]. In China, mortality due to malignant tumor, cerebral vascular disease, heart disease, and respiratory system disease accounts for 80% of all deaths. Furthermore, high out-of-pocket costs for treating these critical illnesses can easily impoverish patients and their households. The economic burden of chronic diseases is as high as 2.6 trillion in China [
5], and more than 40% of 72 million poor people re-entered poverty because of diseases in 2016 [
6]. Healthcare expenses may overburden patients with critical illnesses, which are often financially devastating for their households. With a limited financial protection effect of the BMI system on reducing catastrophic health expenditure (CHE) in China [
7,
8,
9], the economic burden of diseases in Chinese households is higher than generally appreciated, and will continue to increase as the population ages [
3].
To protect urban and rural residents against large medical expenses and relieve medical expenditure burden, the China central government launched critical illness insurance (CII) in August 2012. It was piloted in more than 134 cities from 2013 and implemented nationally in 2016. CII, also called critical illness insurance for urban and rural residents, covers the enrollee of URMI and NRCMS. It reimburses individuals with critical illnesses for high out-of-pocket healthcare payments (OOP) exceeding a certain level after BMI reimbursement [
10]. The central government proposes general guiding principles; for example, the total reimbursement rate is no less than 50% (adjusted to 60% in a 2019 government work report [
11]) when medical bills for necessary treatments after reimbursement under the BMI system exceed annual per capita income levels. All provincial governments have made detailed rules about the financial mechanism, coverage, and reimbursement with respect to CII. This program does not require an additional premium from the insured; it is mainly funded through BMI surpluses. Therefore, it is an extension and supplement to the Chinese BMI system.
Total OOP equaling or exceeding 40% of the household capacity to pay (CTP) is considered catastrophic. This definition for CHE proposed by the WHO is widely accepted [
3,
12,
13]. The CHE is an important indicator for domestic and foreign researchers to evaluate the effects of medical insurance systems. Many studies have assessed CHE worldwide. Barros et al. [
14] estimated catastrophic healthcare expenditure in Brazil using different definitions (including >40% of household CTP), and the results indicated that health insurance coverage did not protect from CHE. A systematic review was undertaken by some researchers in Ghana to evaluate whether enrollment in the National Health Insurance Scheme reduces the likelihood of OOP and CHE. They adopted the thresholds of CHE that OOP for healthcare exceeds 20% of annual household income, 10% of household expenditures, and 40% of subsistence expenditures, respectively. Findings suggested that healthcare costs remain catastrophic for a large proportion of insured households in Ghana [
15]. Two other studies by Xie et al. [
16] and Guo et al. [
17] evaluated the role of the Chinese new rural cooperative medical scheme (NRCMS) on protecting households from CHE using 40% of a household’s capacity to pay as the threshold. Both reported that CHE incidence and intensity decreased after NCMS reimbursement.
Alleviating the economic disease burden and protecting households from CHE were the most important targets of CII [
10]. Although local governments have implemented CII for several years, it is far from achieving universal coverage. For example, approximately 50% of households in Jiangsu Province have not been insured by CII, according to the Sixth National Health Service Survey (NHSS) conducted in Jiangsu Province in 2018. Some studies have researched the implementation effect of CII in China. However, most focused on describing and comparing the design and development of the program policy [
18,
19,
20,
21] and do not provide strong empirical evidence on the topic. In recent years, several studies performed quantitative analyses. Fang et al. [
22] evaluated the effect of CII on the share of medical expenses reimbursed and found that it increases the level of reimbursement to some extent. Zhao et al. [
3] indicated that CII is only partially effective in financial risk protection under total medical expenses rather than covered medical expenses. Another study [
23] focused on the impact of CII on utilization and costs of hospitalization before and after implementation and found that CII can promote inpatient service utilization but increases patients’ economic burden.
All individuals should have access to the health services they need without risk of financial ruin or impoverishment (WHO 2013) [
24]. With China’s aging society, it is necessary to analyze the effect of CII on household CHE and economic burden of diseases. To the authors’ knowledge, due to a short period of time since broad CII implementation and the lack of empirical data, few studies have assessed the effect of CII on CHE. This study used data from the Sixth National Health Service Survey (2018) to calculate the extent of catastrophic health expenditure in urban and rural households in Jiangsu Province and evaluated the effect of critical illness insurance and associated factors on household catastrophic health expenditure. This study clarifies the implementation effect of its critical illness insurance on financial risk protection in Jiangsu Province; it also provides some significant empirical evidence for reforming China’s healthcare insurance.
4. Discussion
Implementing CII is an important step in the Chinese government’s efforts to reduce CHE and alleviate household economic burden due to illness. As China’s population rapidly ages, more residents are suffering from serious diseases, and households are faced with heavy financial burdens. Following nationwide CII implementation, it is essential to study the effect on household financial risks of diseases. Our study used the latest data to calculate the incidence and extent of CHE among rural and urban households and estimate the effect of CII and other factors on CHE incidence and intensity.
Among the 3660 included households in Jiangsu Province, the proportion with any members insured by CII in rural areas was higher than that in urban areas (76.46% vs. 27.35%), which demonstrates the inadequate and imbalanced coverage of CII. The proportion of all sampled households facing CHE at OOP ≥40% of non-food expenditure was 29.18%, which was greater than the proportion (16.5%) reported in another study based on 2015 national data [
29]. In
Table 3, The CHE incidence (Headcount) and intensity (Overshoot) of rural households were significantly higher than those of urban households, irrespective of the threshold used. This result is similar to a study conducted in three cities in eastern China [
30]. Moreover, we found the CHE of rural households was >66.15% of their non-subsistence expenditure (MPO). These findings in regard to CHE incidence (Headcount) and intensity (Overshoot) of households in rural and urban areas indicate that households, especially in rural areas, are still vulnerable to high medical care costs. Lower income level and the absence of social security in rural areas are among the possible causes. Thus, the government should implement more measures focused on rural households to improve the social security system and reduce their CHE.
Regarding the associated factors of CHE in Jiangsu Province, we found that households with a low education householder, lower household income, members with ≥2 chronic diseases, and utilizing outpatient and inpatient services have a higher likelihood of CHE incidence, which is in accordance with other studies [
3,
31,
32,
33,
34]. The above factors are also associated with a higher intensity of CHE. The coefficient of the interaction variable of household members >65 years old and number of household members having CII was −0.006 (
p < 0.05), which indicates that, compared to households without members >65 years old, the number of members covered by CII in households with members >65 years old have stronger negative effect on the intensity of CHE. With regard to CHE incidence, the effect of the interaction variable was significantly positive, which means that households with >65-year-old members are more likely to have CHE when the number of CII-insured members in the household increases. The CII seems to provide protection for households having members >65 years old who are also the insured of CII from high intensity of CHE but fail to reduce the occurrence of CHE among households. We supposed that more CII patients in a household might have a higher likelihood of having expensive diseases in that household. Also, most of the members insured by CII in a household may be the elderly. They are easily suffered from diseases and are vulnerable to higher healthcare expenses. Therefore, it is necessary to further focus policy efforts on elderly and impoverished households, such as increasing welfare subsidies or extending medical assistance programs to enhance financial protection and reduce CHE. Previous reports described significant associations between the type of healthcare facility and CHE [
28,
35]. Our study also found that seeking inpatient services in second-level hospitals or above increased CHE incidence and intensity, which suggests that higher medical costs of high-level hospitals and hospitalization may result in more catastrophic health expenses. In addition, chronic diseases cause both suffering to patients and catastrophic financial burden to households, as demonstrated in several studies [
29,
31,
32,
33,
34,
36]. Therefore, increasing the medical insurance level of patients with chronic diseases and improving chronic disease management could reduce household CHE and provide protection against medical economic risks.
Rural registration had no significant association with CHE incidence, contrary to some existing studies [
29,
30], but had significantly higher CHE intensity compared with urban households. A possible explanation is that there may be non-use of needed healthcare services by patients in rural households because of low income [
37]. However, rural patients utilizing health services tend to go to more expensive high-level hospitals because they offer advanced medical equipment and technology compared to first-level hospitals. Therefore, OOP of rural households may far exceed their non-food expenditure [
38]. The government should take the reimbursement level of rural residents and improvement of first-level hospital services into account. Household size means the number of persons living together in one household. The result of our study that household size was negatively associated with the likelihood of CHE is in agreement with previous reports [
29,
39,
40]. However, several published studies pointed out that large household size is significantly associated with high CHE [
35,
41]. We found a slightly positive correlation between household size and CHE intensity. These findings indicate that larger families protect against CHE incidence but not intensity. As the characteristic result in
Table 1 indicated, rural household size is significantly higher than urban. Larger families are usually found in Chinese rural areas. In China, a nuclear family, consisting of a married couple and their unmarried children, is the mainstream form of family structure in urban areas. While in rural areas, there are more households with three or four generations living together, which are called a stem family or joint family. And generally, the proportion of the elderly in rural households is higher than that in urban areas because the majority of the elderly are employed in agriculture, and they are used to living in rural areas. The young and middle-aged tend to go to cities for better employment and live there. Although larger families may have a good economic ability to reduce the risk of CHE, they are more likely to include more elderly members or children who are susceptible to illness and incur higher costs [
30].
Previous studies concerning the effect of CII on relieving financial burden have yielded mixed results. Some studies held the view that the effect is limited from the perspective of CII reimbursement level [
22,
23,
42], while other studies indicated that economic burden decreased after CII implementation [
3,
43,
44]. In our study, a household with members having CII was a factor to protect households from CHE incidence, but it significantly increased CHE intensity. The number of household members insured by CII did not affect the incidence at a significant level, but this factor significantly decreased the intensity of CHE. A possible explanation is that having CII actually improves households’ capacity to resist economic risk due to critical diseases, and therefore, significantly decreased CHE incidence. However, CII implementation also induced residents’ demands for medical services [
23], which could lead to higher healthcare expenses, even including uncovered medical expenses that are not on the CII list. Therefore, once CHE occurs in a household, the intensity will be greater, especially in lower-income households. As the number of the insured by CII in a household increased, the association of CII with incidence became insignificant. Instead, households with more CII-covered members decreased CHE intensity.
The main purpose of CII is to provide protection against CHE; however, our study showed that this objective has not been fully achieved in Jiangsu Province. In view of the possible causes listed above, policy efforts should focus on the following aspects. First, strategies should be developed to promote the inclusion of more households in CII to effectively decrease household CHE incidence. At the same time, it is necessary for the government to establish a dynamic reimbursement list of CII to expand coverage and reduce uncovered medical expenses. Second, due to the significant effect of the number of CII-covered household members on CHE intensity, the CII coverage rate in Jiangsu province should be continually increased to better relieve economic burden associated with disease. Third, to better exert the protective effect of CII, some medical cost control mechanisms should be established to reduce irrational increases of medical expenses exceeding actual needs. In addition, rural households are in urgent need of policy support because of the higher incidence and intensity of CHE, lower-income, and less social security. Therefore, the government should properly increase the reimbursement rate or reduce the deductible line of CII in rural areas to further alleviate household economic burden of diseases.
Some limitations of this study must be noted here. (1) We calculated household medical expenses, excluding indirect expenses (e.g., transport, food, accommodation costs, etc.) for patients and companions, which is reported to account for a fair proportion of total OOP. This conclusion may lead to an underestimation of CHE. (2) Due to limited access to data, our study used a sub-sample of NHSS (2018) data from Jiangsu Province, which may be not nationally representative. However, Jiangsu Province is a relatively developed area and is typical in the implementation of CII among 31 provinces surveyed. (3) We evaluated the effect of CII on CHE by measuring the incidence and intensity; we did not focus on specific OOP and the actual reimbursement ratio because of a lack of data. Further studies will be done in the future.