Catastrophic Health Expenditure Incidence and Its Equity in China: A Study on the Initial Implementation of the Medical Insurance Integration System

doi:10.21203/rs.2.11043/v1

Download PDF

Research article

Catastrophic Health Expenditure Incidence and Its Equity in China: A Study on the Initial Implementation of the Medical Insurance Integration System

https://doi.org/10.21203/rs.2.11043/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 30 Dec, 2019

Read the published version in BMC Public Health →

You are reading this older preprint version

Read the latest preprint version →

Introduction: By 2013, several regions in China have introduced health insurance integration policies. However, few studies have addressed the impact of medical insurance integration in China. This study investigates catastrophic health expenditure and equity in the incidence of catastrophic health expenditure by addressing its potential determinants in both integrated areas and non-integrated areas in China in 2013. Methods: The primary data are drawn from the fifth China National Health Services Survey in 2013. The final sample comprises 19,788 households (38.4%) from integrated areas and 31,797 households (61.6%) from non-integrated areas. A probit model is employed to decompose the inequality of the incidence of catastrophic health expenditure in line with the methodology used for decomposing the concentration index. Results: The incidence of catastrophic health expenditure in integrated areas is higher than in non-integrated areas (13.87% vs. 13.68%, respectively). The concentration index in integrated areas and non-integrated areas is -0.071 and -0.073, respectively. Average household out-of-pocket health expenditure and average capacity to pay in integrated areas are higher than in non-integrated areas. However, households in integrated areas have lower out-of-pocket expenditures share in capacity to pay than households in non-integrated areas. The majority of the observed inequalities in catastrophic health expenditure can be explained by differences in the health insurance and householders’ educational attainment both in integrated areas and non-integrated areas. However, compared with non-integrated areas, the inequity caused by integrated medical insurance has been significantly reduced in integrated areas. Conclusions: The medical insurance integration system in China is still at the exploring stage; hence, its effects are of limited significance. Regardless of the area, catastrophic health expenditure is associated with pro-poor inequality. Medical insurance, urban-rural disparities, the elderly population, and the use of health services significantly affect the equity of catastrophic health expenditure incidence. Keywords: Catastrophic health expenditure, Medical insurance integration system, Equity, Influencing factors

Health Economics & Outcomes Research

Catastrophic health expenditure

Medical insurance integration system

Equity

Influencing factors

At the end of 1998, China launched a government-run mandatory insurance program, the urban employee basic medical insurance (UEBMI), to replace the previous medical insurance system [1]. In 2003, the new rural cooperative medical system (NRCMS) was launched as a voluntary insurance program to address the vast uninsured rural population [2]. Four years later, the urban residents’ basic medical insurance (URBMI) was developed targeting students, children, and unemployed urban residents not covered by the UEBMI and NRCMS schemes [2]. In 2010, the Chinese government acknowledged three categories in the basic medical insurance system , namely, UEBMI, URBMI, and NRCMS [3]. These insurance schemes have enabled China to achieve near-universal health coverage, with more than 1.3 billion Chinese people (about 97% of the population) having some medical insurance [4].

However, significant urban-rural disparities in health care and social equality have negatively affected the progress towards universal health care. Severe deficiencies exist in the current fragmented medical insurance system, especially in terms of the disparity between urban and rural health care. Li et al. argued that the current NRCMS is of little help to adequately protect the insureds from catastrophic health expenditure (CHE) or poverty caused by diseases [5]. Guo et al. found that the impact of NRCMS in alleviating the financial burden of rural residents is limited, especially among the poor [6]. In 2009, the Chinese government assessed the need to build an urban-rural integrated security management system [7]. Some regions have attempted to integrate the URBMI and NRCMS into the urban-rural resident basic medical insurance (URRBMI) [8]. Five provincial administrative regions’ municipalities (Tianjin and Chongqing, Qinghai province, Ningxia Hui autonomous region, and the Xinjiang Production and Construction Corps), 41 prefecture cities, and 162 counties (districts, county-level cities) had already established the URRBMI at the end of 2011 [9]. In addition, some advanced regions have tried to integrate the three basic medical insurances into a universal social basic medical insurance (USBMI).

As defined in the World Health Report 2000, CHE represents the disruption to material living standards due to large out-of-pocket (OOP) payments for health care in the absence of adequate medical insurance coverage [10]. There is no consensus on the threshold above which health expenditures are considered catastrophic. OA et al. defined CHE as OOP direct medical costs exceeding 10% of monthly household income [11]. Kawabata et al. proposed that health expenditures exceeding 40% of a household’s capacity to pay should be considered catastrophic [12]. The WHO (World Health Organization) defined financial catastrophe as OOP expenditure exceeding 40% of household income net of subsistence needs [13].

Surveys in eighty-nine countries, covering 89% of the world’s population, suggested that 150 million people globally suffer financial catastrophes every year due to OOP medical costs [14]. Developed countries have advanced medical insurance and medical service systems that protect households from catastrophic spending. Choi et al. found that only 3.5% of Korean households suffered from CHE in 2008 (according to the 40% threshold) [15]. However, the incidence of CHE (H_cat) in developing countries is relatively high, especially in low-income developing countries. Ghimire et al. utilized nationally representative data for Nepal and found that the cumulative H_cat is 10.3% per month [16]. A clustered sample survey conducted in Iran in 2008 showed that 11.8% of households faced CHE [17]. In Georgia, the results of a survey showed that 19% of households seeking care had to borrow money or sell personal items to pay for health care, and 16% were unable to afford the medications prescribed [18]. According to a cross-sectional study, in 2008, about 13% of families experienced CHE in China [19].

Previous studies associated several factors to CHE. Household size, the presence of family members over 65 or less than five years old, household members with a chronic disease, residence, hospitalized family members, income, insurance, gender of the household head, and education level, among the others, have been significantly associated to CHE [16, 20-22]. Falconi indicated that poorer, rural, and smaller households, as well as households with older members and individuals with chronic conditions, have larger odds of facing CHE [23]. However, only a few studies have addressed inequality in CHE. Boing et al. found that the poorest households and households headed by the least-educated individuals significantly contribute to increasing social inequality [24]. Moradi et al. showed that income is the most significant determinant of inequality in facing CHE [25]. Wang et al. used a nationally representative dataset and found that household size, per capita income, family members above 65 years of age, and family members with two or more chronic diseases significantly increase CHE inequality [26].

This study’s contribution to the literature is twofold. First, although some areas have implemented a medical insurance integration system in China, only a few studies have evaluated the impact of its implementation. Second, previous studies have identified several potential determinants of equity in the H_cat. However, the existence of factors that can systematically hinder equity also needs to be explored after the implementation of the medical insurance integration system. This study investigates the impact of the current medical integration system on reducing the H_cat and improving equity in H_cat in China while exploring the determinants of inequality in H_cat.

Data sources

The primary data used in this study are drawn from the fifth National Health Services Survey (NHSS, 2013) conducted by the Center for Health Statistics and Information of the Ministry of Health of China. The survey has been carried out every five years since 1993 and collects information regarding the general status of the family, personal status of family members, illness of family members in the two weeks before the survey, hospitalization within one year before the survey, and situation of children under five and women aged 15-64.

Sampling method

The fifth NHSS adopts four-stage, stratified, random sampling, involving 31 administrative divisions (excluding Taiwan, Hong Kong, and Macau) in China. A total of 156 cities are randomly selected in the first stage, representing six geographical locations (eastern cities, eastern rural areas, central cities, central rural areas, western cities, and western rural areas, respectively). In every city, five sampled townships are randomly selected in the next stage, for a total of 780 townships. Among these, 1560 villages are selected in the next stage. Finally, 273688 respondents (from 93613 families) are investigated. The response rate is 82.1% (at the individual level).

In this study, to assess the determinants of inequality in the H_cat in the pioneer integrated areas (IAs) of China, we address pilots that implemented the local medical insurance integration policy before the fifth NHSS. The sample comprises areas that implemented the integration (the integrated pilots). As a result, the sample comprises 33 integrated pilots of 156 sample areas, which belong to 13 administrative divisions. In three regions (Tianjin, Chongqing, and Ningxia), every sample area represents an integrated pilot, and comparable regions (Beijing, Shanxi, and Hebei) by per capita GDP (as of 2012) are used for the analysis. All sample areas of the comparable regions are referenced as non-integrated pilots. In the other 10 regions, some sample areas are integrated pilots, and the remaining sample areas are referenced as non-integrated pilots. Finally, 13 regions (33 integrated pilots) are identified as “IAs”, and the corresponding 13 regions (53 non-integrated pilots) are identified as “non-integrated areas (NIAs)”.

After the data cleaning process, the final sample comprises 19788 households (38.4%) in IAs and 31797 households (61.6%) in NIAs. Quality control has been implemented by supervisors charged with guiding and inspecting every step of the survey. Face-to-face household interviews have been conducted by qualified investigators, and 5% of the sample families have been revisited to check the accuracy of the data (above 95%; otherwise, return visits were made to reinvestigate).

Variables

The relevant characteristics of the head of the household are gender (men, women), marital status (married or other), educational attainment (illiterate, primary school, junior high school, high school and above), occupation (employed, retired, other), and medical insurance (UEBMI, URBMI, NRCMS, USBMI, URRBMI, mix medical insurance, uninsured and other). The relevant household characteristics include household location (eastern/middle/western area), residence (urban or rural), household size (≤2, 3–4, ≥5) and preferred medical institution (primary/non-primary). Household income is ranked in five quintiles according to the WHO, and OOP health expenditure, food consumption expenditure, and medical consumption expenditure at the household level are accounted for. Some household variables are measured as “yes” or “no,” based on the questions “Has any household member been hospitalized in the past year?” “Is there anyone with a chronic disease in the family?” “Is there anyone over 60 years old in the family?” and “Is there anyone under five years old in the family?”

Statistical analysis

CHE is a general term used to describe all types of health expenditures that generate a threat to the capacity to pay (CTP) of a household and its subsistence needs. In this study, CHE is analyzed based on measures recommended by Wagstaff et al., which include the H_cat, CHE gap, CHE inequality, and decomposition of CHE inequality [27]. In this study, the 40% threshold recommended by the WHO is used to calculate CHE [13]. Monthly household consumption expenditure is ranked into quintiles after adjusting it for standard household size. This adjustment allows any differences in health spending across countries to be attributed to factors other than the differential composition of their populations. The poverty line is defined by subsistence spending as the average monthly food expenditure of the household whose food expenditure as a share of total household consumption expenditure falls between the 45^th and 55^th percentiles of the sample. The subsistence spending of each household is calculated as the poverty line multiplied by the standard household size. If a household’s total expenditure is less than this figure, the household is categorized as poor [20]. Household non-subsistence spending is used as a proxy for CTP. However, whenever food expenditure is less than subsistence spending, CTP is defined as total expenditure minus food expenditure.

H_cat describes the proportion of households facing CHE in the sample. In addition to a catastrophic payment headcount, this study adopts a measure analogous to the poverty gap, called “CHE gap”. The mean CHE gap (G_cat) and the mean positive CHE gap (MPG_cat) are used to measure the intensity and severity of CHE, respectively.

The concentration index (CI) is employed to measure the extent of economic inequality in the H_cat. CI is defined as twice the area between the concentration curve and the line of equality and lies between [-1,1] [28]. Its positive value indicates that inequality in the H_cat is more concentrated among the rich, while a value equal to zero indicates that there is no inequality. The larger the absolute value of CI, the greater the H_cat. In line with Wagstaff et al., the W^E_Cat index is calculated to modify the catastrophic payment headcount by the individual rank in the income distribution (which corresponds to the weighted H_cat). [27].

Inequality can be further explained by decomposing the CI into its determinants. Various decomposition methods can quantify each determinant’s contribution to measuring economic inequality while controlling for other determinants. A probit model is employed to decompose the inequality of the H_cat in line with the decomposition method used for the CI. All analyses are performed in Stata version 11.0, and p < .05 is the threshold for statistical significance.

Basic features of sample families

Table 1 reports the summary statistics of the households and the household’s heads’ characteristics. Both in IAs and NIAs, household’s heads are predominantly men, married, with junior high school, employed. The primary insurance scheme for the heads of household is NRCMS (56.2%) in NIAs and URRBMI (56.1%) in IAs. Households in IAs are mainly concentrated in the eastern urban areas. Compared with NIAs, households in IAs are more likely to attend non-primary medical institutions (23.2% vs .18.8%).

Catastrophic health expenditure

At a 40% threshold, the poorest households face the highest proportion of CHE occurrence compared to other quintiles, both in IAs and NIAs; the CI is negative both in IAs and NIAs, thus suggesting that inequality in CHE is biased towards low-income people. The absolute value of the CI is slightly lower for IAs households than NIAs households. Furthermore, the W^E_Cat of IAs households is slightly higher than that of NIAs households (Table 2).

Table 3 presents the average household OOP health expenditure, average household CTP, and the proportion of households affected by poverty and medical impoverishment. In IAs, addressing the income-based groups from the lowest to highest income, the average OOP expenditure ranges from 1041.68 yuan up to 8445.06 yuan, and the OOP share in CTP ranges from 21.90% down to 14.90%. In NIAs, addressing the income-based groups from the lowest to highest income, the average OOP expenditure ranges from 1023.66 yuan up to 8881.48 yuan, and the OOP share in CTP ranges from 21.86% down to 16.08%. Overall, the average household OOP health expenditure and average household CTP in IAs are higher than in NIAs. However, households in IAs face a lower OOP share in CTP than households in NIAs.

Illness behavior and utilization in IAs

“Non-admission rate” is defined as the percentage of patients needing treatment who are not treated in the two weeks before the survey. In IAs, households for which the household head is enrolled in the UEBMI have the highest prevalence rate and non-admission rate in the two weeks before the survey (Figure 1).

Decomposition of inequality of CHE

Table 5 reports the CI and the relative contributions of each determinant of the H_cat inequality in both IAs and NIAs. The β_k coefficient indicates that hospitalized household members, household population aged two or less, family members with chronic diseases, household population equal to three or four members, family members over 60 years of age, living in rural areas, and gender (men) are associated with increased risk of incurring CHE, both in IAs and NIAs. The URRBMI has a positive impact on the CHE occurrence rate among household in IAs.

The third and seventh columns implies the extent to which the respective variable is distributed across wealth. For example, in IAs, the CIs of some determinants such as men, rural, URRBMI, and midland, are negative, meaning that these features are more concentrated among people of lower economic status. In contrast, mixed medical insurance, UEBMI or USBMI, high school and above, employed or retired have a positive CI, thus implying that these features are more concentrated among people of higher economic status.

A positive contribution to socioeconomic inequality means that the considered variable increases inequality. The results of the fifth and last column of Table 5 show that the majority of the observed inequalities in the H_cat in IAs can be attributed to educational attainment (45.47%), medical insurance (29.89%), family members over 60 years of age (19.32%), area (18.55%), and household income (13.38%). Hospitalized members (-42.48%), family members with chronic diseases (-4.85%), and location of residence (-0.39%) are negatively related to CHE inequality, thus implying that these factors reduce CHE inequality in IAs. The total percentage contribution is 102.33%, which means that 2.33% of the positive contribution to inequality in the H_cat is explained by the error term of the regression. In NIAs, the primary positive contribution to inequality is associated with medical insurance (60.73%), educational attainment (33.41%), area (23.88%), household income (23.83%), and family members over 60 years of age (12.99%). Hospitalized family members (-44.00%), preferred medical institution (-12.23%), family members with chronic diseases (-9.64%), occupation (-7.06%), and family members below five years old (-0.60%) are negatively related to CHE inequality in NIAs. The total percentage contribution is 93.77%, which implies that 6.23% of the negative contribution to inequality in the H_cat is explained by the error term of the regression.

This study utilizes nationally representative data to analyze the incidence, intensity, and inequality of H_cat for households in IAs and NIAs after the implementation of the pilot policy of medical insurance integration in China. The results of this study can support decision-makers in formulating policies and relieve the economic burden of disease in vulnerable groups.

The CHEs in both IAs and NIAs in China is calculated. Compared with the previous study, the H_cat has not decreased significantly over the sample period. [20]. The rapid growth in health service demand and medical expenses in China may have played a significant role. The two-week prevalence has increased from 18.86 in 2008 to 24.10 in 2013 [29], and the per capita hospitalization cost has increased from 5234.1 yuan in 2008 to 7858.9 yuan in 2013 [29, 30]. The aging of the Chinese population is also worthy of attention. The proportion of the population aged over 65 in China has risen from 8.3% in 2008 to 9.7% in 2013 [31]. Hence, this study’s results seem to confirm the effect of the current medical reform and medical insurance integration policies.

Both in IAs and NIAs, the poorest families face the lowest OOP expenditure but experience the highest share of OOP payment for health care. This result confirms the findings of a previous study, which argued that low-income families pay a somewhat higher ratio of OOP expenses relative to their household incomes [32]. OOP expenses are higher in the highest income quintile as compared to the lowest income quintile, but households in the highest income quintile suffer a minimal catastrophic impact. This result suggests that although the richest households pay more for health care, they are less likely to suffer a change in their living standards or incur debt due to health care expenses [33]. Furthermore, the proportion of medical impoverishment in IAs is lower than in NIAs. Compared with previous studies, this study’s results show that, for low-income residents, the proportion of medical impoverishment in IAs has significantly decreased [20]. Therefore, the positive impact of the medical insurance integration system on low-income residents is confirmed.

In this study, CI is used to measure inequality in the H_cat. The results indicate that CHE is characterized by inequality concentrated among the poor in both IAs and NIAs. After decomposing the inequality in the H_cat, medical insurance is found to significantly contribute to inequality, especially in NIAs. This result indicates that the medical insurance integration system has improved the equality of H_cat to some extent. Although it has reduced the inequality caused by medical insurance in IAs compared to NIAs, the integrated medical insurance system has failed to solve the inequality in the H_cat caused by medical insurance. In IAs, URRBMI is still at the exploring stage and contribute in favor of the poor. URRBMI adopts "financing by stages, linking payment with treatment" to adapt to the consumption capacity of urban and rural residents characterized by different economic levels, but it also stimulates an invisible inequity, which concentrates on the poor. Low-income residents generally choose a financing level of medical insurance characterized by a low payment threshold and can only benefit from low levels of reimbursement [34]. This phenomenon reflects the heavy medical burden for economically disadvantaged groups, which remains unsolved.

Interestingly, compared to uninsured populations, URRBMI enrollees positively relate to the risk of increasing CHE. Several possible explanations exist for this phenomenon. First, URRBMI enrollees have shown a higher prevalence and visiting rate in the two weeks before the survey compared to the uninsured population. In other words, URRBMI enrollees have greater potential to use health services, which, in turn, increases the risk of CHE. Second, the prevalence of untreated in the URRBMI enrollees is also higher than in the uninsured population. The absence of a doctor’s visit may lead to aggravation of the patient’s condition and lead to CHE. Third, a higher CHE risk may be due to adverse selection in the URRBMI policy. The URRBMI shares the patient's health expenses and reduce the cost burden on families. Thus, people in poor health condition may be more willing to participate in the URRBMI compared to uninsured people, who are generally healthy.

Significant deficiencies still exist in China's medical insurance integration policy. In the implementation of future insurance integration policies, the focus should shift towards the health needs and payment capacity of all classes of citizens. In addition, a reasonable fundraising and payment mechanism need to be established to reduce the inequality caused by medical insurance.

This study’s results suggest that residency positively contributes to inequality in H_cat. In China, due to the dual structure of urban and rural areas, the urban-rural income gap is significant. Previous studies found that the urban-rural income gap accounts for the majority of the national income gap [35, 36]. In addition, the unequal distribution of health resources between urban and rural areas in China further exacerbates the disparities in the health level of urban and rural residents. Wu et al. argued that rural residents are much more financially vulnerable to health crises, and most CHE cases are attributed to rural families [37]. The current medical insurance integration system has only achieved the unified management of urban and rural systems, but gaps still exist between urban and rural residents in their ability to purchase health services. Sun et al. argued that leveling the reimbursement ratios between urban and rural residents is needed for achieving health equality [9]. Therefore, after the integration of the urban and rural medical insurance system, the equality of the financing burden for rural residents should be addressed. In areas with large urban-rural gaps, “one system, two files” or “one system and multiple files” can be implemented, allowing rural residents to choose between various grades, and transition to “one system and one file” may be pursued when appropriate. In addition, the government needs to invest more funds to further expand the Social Medical Insurance Programs for rural low-income people to avoid CHE.

The education level of the household head contributes to pro-poor inequality in the H_cat. This may be due to the relatively poor health care awareness of the heads of households with lower levels of education, which, in turn, makes them more likely to incur in CHE. However, in IAs, the contribution rate of households’ income factors to CHE inequality is smaller than in NIAs. Medical integration policy designers should recognize the need to narrow income disparities. Further narrowing the income gap and providing fair access to education are aspects that cannot be ignored in the development of social security systems.

The presence of family members over 65 years old is the primary contributor to CHE inequality. This pro-poor contribution indicates that low-income elderly households are more likely to experience CHE. Previous studies showed that the presence of family members over 65 years of age increases the OOP health expenditures as this category of the population is vulnerable to diseases and health dysfunctional [38]. Although the current reimbursement rate for medical insurance for the elderly is continuously increasing, the costs of nursing care, transportation, and nutrition due to illnesses are not covered by medical reimbursement. The problem of aging in China has become severe. Older people (aged 60 or older) are expected to outnumber people between 0 and 14 years old by 2020 [39]. In addition, the People’s Republic of China’s one-child policy increases the pressure on home care for the elderly. Zhang et al. found that the current medical insurance does not play a significant role in reducing inequality among patients who need long-term care in China [40]. Therefore, the reform of the medical insurance system, in addition to integrating the existing medical insurance system, should also consider introducing a medical insurance system for the elderly and covering long-term care services.

Furthermore, the hospitalization of a family member is more likely to occur in wealthier households. This phenomenon reduces inequality in CHE, disfavoring the rich. In other words, as poor people use fewer inpatient care, they are less affected by the catastrophic impact of spending on such services. As the use of inpatient services is concentrated in wealthier families, this phenomenon increases the chance of CHE in such families, thus reducing inequality in the number of families facing CHE in different socio-economic groups. Although hospitalization reduces the inequality of CHE occurrences, hospitalization is also positively associated with CHE occurrences. Deng et al. found that differentiation in copayment design can influence patients’ medical-care-behavior in the Chinese tiered health care system [41]. In the future reform process, the Chinese government should focus on the combination of the tiered health care system and the medical insurance integrated system to reduce unnecessary health expenditures of patients and ultimately reduce the H_cat.

This study suffered several limitations; hence, the results should be interpreted with great caution. First, the data used for analysis reflect the initial results of the integration policy, but the implementation of the integration process requires long-term observation and evaluation. Second, poor households may refuse treatment due to their inability to afford medical expenses, which may lead to lower than actual estimates of the expenditure for CHE. We plan to keep collecting relevant data from the areas of medical insurance integration and compare new data with the results of this study to further analyze the implementation effect of China's medical insurance integration policy.

Conclusions

As the medical insurance integration policy is still at the exploring stage, its effect is still of limited significance. Both in IAs or NIAs, CHE shows pro-poor inequality. Medical insurance, urban-rural disparities, the growing number of the elderly, and the use of health services affect inequality in the H_cat. In the implementation of future insurance integration policies, the Chinese government should focus on the utilization of health services and the gap between urban and rural areas, establishing a reasonable fundraising and payment mechanism. This study’s results also suggest that the Chinese government should consider introducing a medical insurance system for the elderly and covering long-term care services.

UEMIB: urban employee basic medical insurance; NRCMS: new rural cooperative medical system; URBMI: urban residents’ basic medical insurance; CHE: catastrophic health expenditure; URRBMI: urban-rural resident basic medical insurance; USBMI: universal social basic medical; OOP: out-of-Pocket health payments; WHO: World Health Organization; H_cat: catastrophic health expenditure incidence; NHSS: National Health Services Survey; IA: integrated area; NIA: non-integrated area; CTP: capacity to pay; G_cat: the mean gap of catastrophic health expenditure; MPG_cat: the mean positive gap of catastrophic health expenditure; W^E_Cat: the weighted H_cat; CI: concentration index.

Ethics approval and consent to participate

Ethical approval for this investigation was obtained from the Medical Ethics Committee, Harbin Medical University.

Consent for publication

Not applicable.

Availability of data and material

Please contact corresponding author for data requests.

Competing interests

The authors declare that they have no competing interests.

Funding

This work was supported by the National Natural Science Foundation of China (Grant No. 71333003, 71804036) and Natural Science Foundation of Heilongjiang Province, China (Grant No. YQ2019G003).

Authors’ contributions

HL conducted calculations, analyzed results and drafted the manuscript. QHW and NN were responsible for the overall design of the research, organized and conducted the survey. JHW, XYQ, MMZ and LHS designed the analyses framework. LJG, ZK and MLJ revised the paper. LP, RHC, BHL assisted with the literature review and data collection. All authors approval of the current version of this manuscript for publications.

Acknowledgements

We are thankful to Cui Li for her help in analyzing the data and guiding the writing of manuscript.

1. Huang F, Gan L: The Impacts of China's Urban Employee Basic Medical Insurance on Healthcare Expenditures and Health Outcomes. Health Econ 2017, 26(2):149-163.

2. He AJ, Wu S: Towards Universal Health Coverage via Social Health Insurance in China: Systemic Fragmentation, Reform Imperatives, and Policy Alternatives. Appl Health Econ Health Policy 2017, 15(6):707-716.

3. The Central People’s Government of the People’s Republic of China. People's Republic of China Social Insurance Law. 2010 (Presidential Decree No. 35). www.gov.cn/zxft/ft215/content_ 1748773.htm. Accessed 13 Aug 2018.

4. WM Yin: To accelerate the development of people’s health and promote the innovation of medical insurance system. In: People’s Daily. 2016.

5. Li Y, Wu Q, Liu C, Kang Z, Xie X, Yin H, Jiao M, Liu G, Hao Y, Ning N: Catastrophic health expenditure and rural household impoverishment in China: what role does the new cooperative health insurance scheme play? PLoS One 2014, 9(4):e93253.

6. Guo N, Iversen T, Lu M, Wang J, Shi L: Does the new cooperative medical scheme reduce inequality in catastrophic health expenditure in rural China? BMC Health Services Research 2016, 16(1).

7. The Central People’s Government of the People’s Republic of China. The Central Committee of the Communist Party of China Opinions' on Deepening the Reform of the Medical and Health System. 2009.

8. The Central People’s Government of the People’s Republic of China.State Council comment on the integration of urban and rural residents’ basic medical insurance system. In.: The Central People’s Government of the People’s Republic of China; 2016.

9. Ma J, Chao, Song, Ze, Gu, Hai: Inequality of Opportunity in Health Care in China: Suggestion on the Construction of the Urban-Rural Integrated Medical Insurance System. Mpra Paper 2013.

10. World Health Organization: The world health report 2000 - Health systems: improving performance. Geneva. In. WHO; 2000. https://www.who.int/whr/2000/en/. Accessed 20 Aug 2018.

11. O'Donnell O, Doorslaer EV, Wagstaff A, Lindelow M, O'Donnell O, Doorslaer EV, Wagstaff A, Lindelow M: Analyzing health equity using household survey data: a guide to techniques and their implementation: The World Bank; 2008.

12. Kei Kawabata, Ke Xu, Guy Carrin: Preventing impoverishment through protection against catastrophic health expenditure. Bulletin of the World Health Organization 2002, 80(8):612.

13. Ke Xu: Distribution of health payments and catastrophic expenditures methodology. In.: Geneva: Department of Health System Financing, World Health Organization; 2005.

14. Xu K, Evans DB, Carrin G, Aguilar-Rivera AM, Musgrove P, Evans T: Protecting households from catastrophic health spending. Health Aff (Millwood) 2007, 26(4):972-983.

15. Choi JW, Choi JW, Kim JH, Yoo KB, Park EC: Association between chronic disease and catastrophic health expenditure in Korea. BMC Health Serv Res 2015, 15:26.

16. Ghimire M, Ayer R, Kondo M: Cumulative incidence, distribution, and determinants of catastrophic health expenditure in Nepal: results from the living standards survey. Int J Equity Health 2018, 17(1):23.

17. Kavosi Z, Rashidian A, Pourreza A, Majdzadeh R, Pourmalek F, Hosseinpour AR, Mohammad K, Arab M: Inequality in household catastrophic health care expenditure in a low-income society of Iran. Health Policy Plan 2012, 27(7):613-623.

18. Skarbinski J, Walker HK, Baker LC, Kobaladze A, Kirtava Z, Raffin TA: The Burden of Out-of-Pocket Payments for Health Care in Tbilisi, Republicof Georgia. Jama 2002, 287(8):1043.

19. Wu QH LY, Xu L, Hao YH.: Effect of health insurance on reduction of catastrophic health expenditure in China. Chinese J Health Policy 2012, 05(9):62-66.

20. Li Y, Wu Q, Xu L, Legge D, Hao Y, Gao L, Ning N, Wan G: Factors affecting catastrophic health expenditure and impoverishment from medical expenses in China: policy implications of universal health insurance. Bull World Health Organ 2012, 90(9):664-671.

21. Buigut S, Ettarh R, Amendah DD: Catastrophic health expenditure and its determinants in Kenya slum communities. Int J Equity Health 2015, 14:46.

22. Aregbeshola BS, Khan SM: Determinants of catastrophic health expenditure in Nigeria. The European Journal of Health Economics 2017, 19(4):521-532.

23. Proaño Falconi D, Bernabé E: Determinants of catastrophic healthcare expenditure in Peru. International Journal of Health Economics and Management 2018.

24. Boing AC, Bertoldi AD, Barros AJDd, Posenato LG, Peres KG: Socioeconomic inequality in catastrophic health expenditure in Brazil. Revista de Saúde Pública 2014, 48(4):632-641.

25. Moradi T, Naghdi S, Brown H, Ghiasvand H, Mobinizadeh M: Decomposing inequality in financial protection situation in Iran after implementing the health reform plan: What does the evidence show based on national survey of households' budget? Int J Health Plann Manage 2018.

26. Wang Z, Li X, Chen M: Catastrophic health expenditures and its inequality in elderly households with chronic disease patients in China. Int J Equity Health 2015, 14:8.

27. Doorslaer AWEv: Catastrophe and Impoverishment in Paying for Health Care: With Applications to Vietnam 1993-98. Health Economics 2003, 12(11):921-933.

28. Wagstaff A: The bounds of the concentration index when the variable of interest is binary, with an application to immunization inequality. Health Economics 2010, 14(4):429-432.

29. National Health and Family Planning Commission: China health statistics yearbook 2013. Beijing Union Medical University Press 2013:126-251.

30. National Health and Family Planning Commission: China health statistics yearbook 2014. Beijing Union Medical University Press 2014:125.

31. National Bureau of Statistics: China statistical yearbook 2018. China Statistical Publishing House 2018.

32. Waters HR, Anderson GF, Mays J: Measuring financial protection in health in the United States. Health Policy 2004, 69(3):339-349.

33. Loganathan T, Lee WS, Lee KF, Jit M, Ng CW: Household catastrophic healthcare expenditure and impoverishment due to rotavirus gastroenteritis requiring hospitalization in Malaysia. PLoS One 2015, 10(5):e0125878.

34. Qiu Yulin WZ: The Evaluation and Analysis of Integrating Urban and Rural Residents Basic Medical Insurance System. China Health Insurance 2018(2):16-20.

35. Wan G: UNDERSTANDING REGIONAL POVERTY AND INEQUALITY TRENDS IN CHINA: METHODOLOGICAL ISSUES AND EMPIRICAL FINDINGS. Review of Income & Wealth 2010, 53(1):25-34.

36. Chen J, Fang F, Hou W, Li F, Pu M, Song M: Chinese Gini Coefficient from 2005 to 2012, Based on 20 Grouped Income Data Sets of Urban and Rural Residents. Journal of Applied Mathematics,2015,(2015-3-2) 2015, 2015(1):1-16.

37. Wu D, Yu F, Nie W: Improvement of the reduction in catastrophic health expenditure in China's public health insurance. PLoS One 2018, 13(4):e0194915.

38. Ebaidalla EM, Ali MEM: Determinants and Impact of Households’s Out-of-Pocket Health Care Expenditure in Sudan: Evidence From Urban and Rural Population. Working Papers 2017.

39. Byles J, Curryer C, Edwards N, Weaver N, D'Este C, Hall J, Kowal P: The Health of Older People in Selected Countries of the Western Pacific Region; 2014.

40. Zhang Z, Wang J, Jin M, Li M, Zhou L, Jing F, Chen K: Can medical insurance coverage reduce disparities of income in elderly patients requiring long-term care? The case of the People's Republic of China. Clin Interv Aging 2014, 9:771-777.

41. Deng Y, Li N, Xu X, Jiang Z: Differentiation copayment design to improve the multi-tiered healthcare system in China. In: Automation Science and Engineering: 2018; 2018: 201-206.

Table 1. Characteristics of households and household heads.

Variable	Total N (%)	Integrated area n (%)	Non-integrated area n (%)
*Household head*
Gender
Men	38221 (74.1)	14706 (74.3)	23515 (74.0)
Women	13364 (25.9)	5082 (25.7)	8282 (26.0)
Marital status
Unmarried and other	8201 (15.9)	3339 (16.9)	4862 (15.3)
Married	43384 (84.1)	16449 (83.1)	26935 (84.7)
Educational attainment
Illiterate	5332 (10.3)	2052 (10.4)	3280 (10.3)
Primary school	14594 (28.3)	5553 (28.1)	9041 (28.4)
Junior high school	18050 (35.0)	6740 (34.1)	11310 (35.6)
High school and above	13609 (26.4)	5443 (27.5)	8166 (25.7)
Occupation
Employed	35080 (68.0)	13074 (66.1)	22006 (69.2)
Retired	9720 (18.8)	3844 (19.4)	5876 (18.5)
Other	6785 (13.2)	2870 (14.5)	3915 (12.3)
Medical insurance
UEBMI	14697 (28.5)	6171 (31.2)	8526 (26.8)
URBMI	2715 (5.3)	-	2715 (8.5)
NRCMS	17875 (34.7)	-	17875 (56.2)
USBMI	930 (1.8)	930 (4.7)	-
URRBMI	11104 (21.5)	11104 (56.1)	-
Mixed medical insurance	3050 (5.9)	1106 (5.6)	1944 (6.1)
Uninsured and other	1214 (2.4)	477 (2.4)	737 (2.3)
*Household*
Location
Eastern	20401 (39.5)	10791 (54.5)	9610 (30.2)
Middle	14384 (27.9)	2397 (12.1)	11987 (37.7)
Western	16800 (32.6)	6600 (33.4)	10200 (32.1)
Residence
Urban	26382 (51.5)	10801 (54.6)	15581 (49.0)
Rural	25203 (48.9)	8987 (45.4)	16216 (51.0)
Household size
≤2	23459 (45.5)	8882 (44.9)	14577 (45.8)
3-4	21351 (41.4)	8077 (40.8)	13274 (41.7)
≥5	6775 (13.1)	2829 (14.3)	3946 (12.4)
Hospitalized member
Yes	11147 (21.6)	3987 (20.1)	7160 (22.5)
No	40438 (78.4)	15801 (79.9)	24637 (77.5)
Member with chronic disease
Yes	23634 (45.8)	9342 (47.2)	14292 (44.9)
No	27951 (54.2)	10446 (52.8)	17505 (55.1)
Member > 60 years of age
Yes	23367 (45.3)	9030 (45.6)	14337 (45.1)
No	28218 (54.7)	10758 (54.4)	17460 (54.9)
Member < 5 years of age
Yes	8712 (16.9)	3210 (16.2)	5502 (17.3)
No	42873 (83.1)	16578 (83.8)	26295 (82.7)
Preferred medical institution
Primary medical institutions	41028 (79.5)	15193 (76.8)	25835 (81.2)
Non-primary medical institutions	10557 (20.5)	4595 (23.2)	5962 (18.8)
Household income
Quintile I (Poorest)	10328 (20.0)	3962 (20.0)	6366 (20.0)
Quintile Ⅱ	10329 (20.0)	3954 (20.0)	6375 (20.0)
Quintile Ⅲ	10298 (20.0)	3957 (20.0)	6341 (19.9)
Quintile Ⅳ	10345 (20.1)	3958 (20.0)	6387 (20.1)
Quintile Ⅴ (Richest)	10285 (19.9)	3957 (20.0)	6328 (19.9)

UEBMI: urban employee basic medical insurance; URBMI: urban resident basic medical insurance; NRCMS: new rural cooperative medical scheme; URRBMI: urban and rural residents’ basic medical insurance system. USBMI: universal social basic medical insurance.

Table 2. H_cat, CHE gap, CI, and WE Cat at the 40% threshold.

	Integrated areas	Non-integrated areas
H_cat (%)
Quintile I (Poorest)	17.19	17.45
Quintile Ⅱ	15.33	13.95
Quintile Ⅲ	12.82	12.46
Quintile Ⅳ	11.99	11.55
Quintile Ⅴ (Richest)	11.98	12.99
Total	13.87	13.68
G_cat(%)	2.67	2.61
MPG_cat(%)	19.25	19.08
CI	-0.071	-0.073
WE Cat (%)	14.85	14.68

H_cat: catastrophic health expenditure incidence; G_cat: the mean catastrophic health expenditure gap; MPG_cat: the mean positive catastrophic health expenditure gap; CI: concentration index; WE Cat: the weighted catastrophic health expenditure incidence.

Table 3. Summary statistics for OOP, CTP, OOP/CTP, and medical impoverishment.

	Integrated areas					Non-integrated areas
Household income	Quintile Ⅰ	Quintile Ⅱ	Quintile Ⅲ	Quintile Ⅳ	Quintile Ⅴ	Quintile Ⅰ	Quintile Ⅱ	Quintile Ⅲ	Quintile Ⅳ	Quintile Ⅴ
OOP (Yuan)	1144.72	2225.46	3141.78	4415.60	9454.39	974.18	1934.9	2743.83	3945.66	8255.18
Average CTP (Yuan)	5637.43	12319.50	19411.68	29691.08	65956.65	4794.29	10302.79	16255.69	24766.32	50333.34
OOP/CTP (%)	21.60	19.06	17.11	15.74	14.51	22.09	19.65	17.56	16.37	15.83
Medical impoverishment (%)	6.46	13.10	5.56	1.82	1.77	7.60	15.40	5.25	1.96	1.30

OOP: out-of-pocket; CTP: capacity to pay.

Table 4. The disintegration of H_catinequality in integrated areas and non-integrated areas.

Integrated areas				Non-integrated areas
Variable	β	CI	Contribution (%)	Variable	β	CI	Contribution (%)
Hospitalized member ^***			-42.48	Hospitalized member^***			-44.00
Yes	0.946	0.096		Yes	0.881	0.099
No	Reference			No	Reference
Chronic disease member ^***			-4.85	Chronic disease member^***			-9.64
Yes	0.396	0.015		Yes	0.434	0.029
No	Reference			No	Reference
Member > 60 years of age ^***			19.32	Member > 60 years of age ^***			12.99
Yes	0.259	-0.098		Yes	0.233	-0.073
No	Reference			No	Reference
Member < 5 years of age		-0.014	0.15	Member < 5 years of age			-0.60
Yes	0.039			Yes	-0.042	-0.050
No	Reference			No	Reference
Preferred medical institution			3.46	Preferred medical institution ***			-12.23
Primary medical institutions	0.040	-0.070		Primary medical institutions	-0.113	-0.076
Non-primary medical institutions	Reference			Non-primary medical institutions	Reference
Household income			13.38	Household income			23.83
Quintile I (Poorest)	Reference			Quintile I (Poorest)	Reference
Quintile Ⅱ	0.032	-0.400		Quintile Ⅱ	-0.037	-0.399
Quintile Ⅲ	-0.035	0.000		Quintile Ⅲ ^**	-0.088	0.001
Quintile Ⅳ	-0.067	0.400		Quintile Ⅳ ^**	-0.113	0.401
Quintile Ⅴ (Richest)	-0.002	0.800		Quintile Ⅴ (Richest)	-0.054	0.801
Household size			7.01	Household size			1.57
≤2 ^***	0.788	-0.026		≤2 ^***	0.646	-0.019
3-4 ^***	0.347	0.038		3-4 ^***	0.294	0.039
≥5	Reference			≥5	Reference
Sex			1.22	Sex			2.70
Men ^***	0.095	-0.012		Men ^***	0.120	-0.019
Women	Reference			Women	Reference
Marital status			2.34	Marital status			3.90
Unmarried and other ^**	-0.066	0.025		Unmarried and other ^**	-0.091	0.029
Married	Reference			Married	Reference
Educational attainment			45.47	Educational attainment			33.41
Illiterate	Reference			Illiterate	Reference
Primary school ^***	-0.180	-0.181		Primary school ^***	-0.213	-0.183
Junior high school ^***	-0.346	-0.003		Junior high school ^***	-0.321	0.007
High school and above ^***	-0.486	0.317		High school and above ^***	-0.401	0.327
Occupation			9.16	Occupation			-7.06
Employed ^***	-0.367	0.006		Employed ^***	-0.394	-0.029
Retired ^**	-0.139	0.156		Retired ^*	-0.091	0.283
Other	Reference			Other	Reference
Location of residences			-0.39	Location of residences			4.30
Eastern ^***	0.045	0.045		Eastern ^***	-0.109	0.080
Middle ^***	-0.126	-0.126		Middle ^***	-0.186	0.000
Western	Reference			Western	Reference
Residence			18.55	Residence			23.88
Urban	Reference			Urban	Reference
Rural ^***	0.127	-0.194		Rural ^***	0.148	-0.189
Medical insurance			29.89	Medical insurance			60.73
UEBMI	0.117	0.232		UEBMI ^***	-0.262	0.354
USBMI	-0.185	0.303		URBMI	-0.006	-0.016
URRBMI ^*	0.205	-0.191		NRCMS	0.105	-0.195
Mixed medical insurance	-0.165	0.357		Mixed medical insurance	-0.140	0.252
Uninsured and other	Reference			Uninsured and other	Reference

CI: concentration index; UEBMI: Urban Employee Basic Medical Insurance; URBMI: Urban Resident Basic Medical Insurance; NRCMS: New Rural Cooperative Medical Scheme; URRBMI: urban and rural residents’ basic medical insurance system. USBMI: Universal Social Basic Medical Insurance.

* p < 0.05; ** p < 0.01; *** p < 0.001

Download PDF

Journal Publication

published 30 Dec, 2019

Read the published version in BMC Public Health →

Editorial decision: Major revision
17 Sep, 2019
Review #2 received at journal
14 Sep, 2019
Review #1 received at journal
15 Aug, 2019
Reviewer #2 agreed at journal
14 Aug, 2019
Reviewers invited by journal
18 Jul, 2019
Reviewer #1 agreed at journal
18 Jul, 2019
Submission checks completed at journal
03 Jul, 2019
Editor assigned by journal
27 Jun, 2019
Editor invited by journal
26 Jun, 2019
First submitted to journal
25 Jun, 2019

You are reading this older preprint version

Read the latest preprint version →

Catastrophic Health Expenditure Incidence and Its Equity in China: A Study on the Initial Implementation of the Medical Insurance Integration System

Status:

Journal Publication

Version 1

Abstract

Figures

Introduction

Methods

Data sources

Sampling method

Variables

Statistical analysis

Results

Basic features of sample families

Catastrophic health expenditure

Illness behavior and utilization in IAs

Discussion

Conclusions

List of abbreviations

Declarations

Ethics approval and consent to participate

Consent for publication

Availability of data and material

Competing interests

Funding

Authors’ contributions

Acknowledgements

References

Tables

Status:

Journal Publication

Version 1