Background

In 2009, China carried out the health-care reform, proposing the near-term goal of “effectively reducing residents’ burden of medical expenses and effectively alleviating ‘difficult and expensive to see a doctor’“, and the long-term goal of “establishing a sound basic medical and health system covering urban and rural residents, and providing the people with safe, effective, convenient, and inexpensive medical and health services”. Health-care reform has undergone major social and economic reforms, from pilot to comprehensive promotion, from industry reform to Healthy China strategy, from rapid population aging to major adjustments in family planning policies, which have brought fundamental changes to peoples’ lives and the society [1].

Since the implementation of the poverty alleviation policy, China has made great achievements. The number of poor people in the country had decreased from 165.67 million in 2010 to 30.46 million in 2017. The incidence of poverty decreased from 17. 2% in 2010 to 4.5% in 2016 [2]. However, among the current remaining poor people in China, poverty due to illness and return to poverty due to illness are the most prominent causes of poverty. Among the rural poor in my country, the proportion of poverty-stricken by disease and return to poverty due to illness reached 44.1% [3]. The incidence of disease has been one of the main causes of poverty, leaving patients and their families falling into short-term or long-term poverty [4].

CHE is a general term used to describe various health care expenditures that threaten households’ financial conditions to maintain their survival needs [5]. WHO defined CHE as health expenditures being more than 40% of household’s capacity to pay (CTP) [6, 7]. In the past decade, great achievements have been made in the National Health System Reform (NHSR) with significant improvement of access, quality and equity of health services, but there are also challenges that cannot be ignored, such as the roaring health-care costs. This is especially true in western China where the per capital income is far behind eastern China [8]. On the other side of the coin, improvement of access to health services, especially hospital services, may cause households to pay a large proportion of effective earnings, pushing households into financial hardship even poverty [9, 10].

Even though China has almost built up a universal health care system, which has played an important role in reducing the economic burden of residents’ diseases, the current basic medical security system is still sub-optimal to prevent residents’ CHE and poverty due to illness. The previous payment method of health expenditure in China was fee-for-service (FFS), which is one of the main reasons for the excessively high medical and health expenditure [11] . A study showed that in 2003, 2008 and 2013, 13.6, 15.1 and 13.8% of households experienced CHE, with a impoverishment rate of 8.6% due to disease, and 64% of non-survival expenditures for households actually incurring CHE, suggesting that CHE for the population had not improved or even increased [12], being heavy financial burden on the sick households and the society, and affected the health-care affordability of the population [13]. CHE is an important indicator of the economic incidence of disease, affecting the achievement of the “Universal Health Coverage” and the equity of the health financing system, as well as on social equity. What’s more, CHE is relative to each household, and there may be a significant imbalance between the income groups in terms of the degree of health financial security, the utilization of medical resources and the tolerance of disease incidence [14].

Therefore, the objective of this study was to examine the CHE in mainland China in the last two decades and its association with socioeconomic status of residents. And this research also explored the trends of CHE in the last two decades and gaps between regions.

Methods

A systematic review was conducted according to the standard of preferred reporting items for systematic reviews and meta-analyses (PRISMA) [15].

Literature search

We systematically searched Chinese literature databases including China Journal Full Text.

Database, Chinese Biomedical Journal Database, Wan fang Data Resource System, VIP Database, and English literature databases including PubMed, SCI, Embase and Cochrane library. The search terms included “catastrophic health expenditure,” “catastrophic medical expenditure,” “poverty-causing health expenditure,” “impoverishment,” “poverty-induced poverty,” “return to poverty due to illness,” “catastrophic health spending,” “catastrophic health spending,” “health payments,” “death by diseases,” “return to poverty due to illness.” The search strategy was firstly formed in PubMed and adopted for other databases (see appendix search strategy). In addition, reference lists of included studies were scanned for more eligible studies. The search period was from January 2000 to June 2020.

Inclusion criteria

The inclusion criteria were: (1) research in mainland China; (2) quantitative studies including randomized controlled trial, non-randomized controlled trial, controlled before-after study, non-controlled before after study, cross-sectional study, cohort study or interrupted time series study; (3) the threshold for the incidence of CHE is 40%.; (4) all study populations; (5) and reported the incidence of CHE and economic status of participants. The language of literature was limited to Chinese and English, since almost all research in China were published in these two languages.

Study selection

Firstly, two reviewers (Qingqing Yuan and Yuxuan Wu) screened the titles and abstracts of the identified citations and removed those obviously not relevant to the subject. Then, according to the inclusion criteria, two reviewers (Yuxuan Wu and Furong Li) select eligible studies by reading the full text. IF there was a disagreement, the three discussed until a consensus was reached. If consensus could not be reached, a senior researcher (Kun Zou orDandi Chen) was consulted for final decision.

Data extraction and quality evaluation

Data Extraction and quality evaluation was performed by two reviewers independently (Qingqing Yuan, Yuxuan Wu). Different opinions were decided by discussion.

Data were collected using pre-designed data extraction table, including the author, year of publication, study location, study period, research design, characteristics of participants, sample size, number of CHE and economic status. If there were multiple reports on the harmonized data, priority is given to the most comprehensive publication. If two survey subjects come from the same survey data, select the literature with the most complete data to be included in the study. The quality of included studies was evaluated using the AHRQ scale [16]. The AHRQ had 11 entries, with an answer of “yes” with 1 point, and an answer of “no” or “not clear” with 0 point, with a total score of 0 to 11.

Statistical analysis

Firstly, heterogeneity between included studies was examined using I2, with I2 > 50% indicated substantial heterogeneity. The pooled incidence of CHE was estimated with 95% confidence interval (CI) using fixed effect model if there was no heterogeneity (I2 < 50%). Otherwise, random effect model was used. Subgroup analysis was conducted by study period (three-years interval from 2006 to 2020), residency (urban or rural areas), the level of socio-economic development (categorized into east, central, and western regions, if the object of the survey area has multiple cities, then divided into other. The level of social and economic development in the east is the best, in the middle is general, and in the west is poor) [17], family income levels (poverty, low-income levels, middle-income, high-income, the poverty standard is based on the poverty line or the minimum living security line stipulated by the local civil affairs department, then divide the remaining families into three equal groups according to their income) [18], to explore their relationships with CHE. We also compared the growth relationship between the incidence of CHE and total health expenditure per capita (data from the National Bureau of Statistics of China) in different time periods. Meta-regression was performed to the disease type, the secular trend, the rural and urban residency, the social development level by region and the household income level adjusted for each other. All statistical analysis was conducted used Stata 15.0 software.

Results

Study selection

A total of 4874 citations were retrieved, 575 were likely to be relevant after viewing the titles and abstracts. Then by reading the full texts, 47 eligible studies (44 in Chinese and 3 in English) with 151,911 participants were finally included. (Fig. 1).

Fig. 1
figure 1

PRISMA flowchart of study selection

Full size image

Characteristics of included studies

A total of 47 cross-sectional studies including 151,911 participants were included in this study. The locations of studies were across-provinces (17), Yunnan (6), Hubei (4), Shandong, (4), Shanghai (3), Sichuan (3), Hebei (3), Shaanxi (2), Gansu (2), Jiangsu (2),Chongqing (2), Shanxi (2), Guangdong (1), Heilongjiang (1), Jiangxi (1), Liaoning (1), Qinghai (1), and Xingjiang (1). The included participants were general population (40), middle-aged and elderly people (4), rural villagers (30), cancer patients (2), HBV patients (2), etc. The quality scores of most of studies were between 4 and 9 points (81.67%) (Table 1).

Table 1 Characteristics of included studies
Full size table

Pooled incidence of catastrophic health expenditure for Chinese residents

The overall incidence of CHE in the past 14 years (2006 to 2020) among Chinese residents was 23.3% (95% CI: 21.1 to 25.6%), and there was a significant heterogeneity between the studies (I2 = 99.3%, P < 0.001), so random effect model was used (Fig. 2).

Fig. 2
figure 2

Meta-analysis of the combined incidence of catastrophic health expenditures for Chinese residents

Full size image

1

Subgroup analysis

Secular trend

In general, with the change of time, the incidence of CHE showed an upward trend. It was 18.6% (95%CI: − 1.5 to 38.8%), 19.6% (95%CI:18.2 to 21.0%), 18.0% (95%CI:14.9 to 21.0%), 27.8% (95%CI:22.6 to 33.0%), and 25.5% (95%CI:18.8 to 32.2%) for year 2006 to 2008, 2009 to 2011, 2012 to 2014, 2017 to 2017, and 2018 to 2020 respectively (Fig. 3, Additional file 1: figure 1 for forest map).

Fig. 3
figure 3

The time trends of catastrophic health expenditures and per capital totatl health expenditure for Chinese residents

Full size image

Rural and urban residency

The incidence of CHE in rural area was 25.0% (95%CI: 20.9 to 29.1%), higher than the urban 20.9% (95%CI:18.3 to 23.4%) (Additional file 1: figure 2 for forest map).

Social development level by region

The incidence of CHE in the central was higher than in the eastern, west region. The pooled incidence of CHE was 25.4% (95%CI: 18.4 to 32.3%) in the central, 25.0% (95%CI: 19.2 to 30.8%) in eastern China, 23.1% (95%CI: 17.9 to 28.2%) in the west and 18.8% (95%CI: 16.5 to 21.2%) in the rest (Additional file 1: figure 3 for forest map).

Household income level

The incidence of CHE in poverty group is higher than that of low-income group, middle-income group and high-income group. The incidence of CHE in poverty households was 30.9% (95%CI:22.4 to 39.5%), For the low-income, middle-income and high-income households, the CHE incidence was 20.3% (95% CI: 17.0 to 23.6%), 19.9% (95% CI: 15.6 to 24.1%), and 23.7% (95% CI: 18.0 to 29.3%), respectively (Additional file 1: figure 4 for forest map).

Meta-regression of associated factors of incidence of catastrophic health expenditure

In all studies, a meta-regression analysis was conducted to explore the impact of heterogeneous sources (including disease type, secular trend, rural and urban residency, social development level by region and household income level). The results show the disease type (p > 0.05), the secular trend (p > 0.05), the rural and urban residency (p > 0.05) and the social development level by region (p > 0.05) and the household income level (p > 0.05) are non-significant regulators (Table 2).

Table 2 The time trends of catastrophic health expenditures for Chinese residents
Full size table

Publication bias

There was statistically significant publication bias in included studies (p<0.001).

The funnel plot of publication bias was shown in Additional file 1: Figure 5.

Discussion

There are four main findings in this study. First, the incidence of CHE in rural areas was significantly higher than that in urban areas. Second, the incidence of CHE in central areas was higher than that in eastern and western areas, and lowest in the other areas. Third, poverty group had higher CHE incidence than that of low-income group, middle-income group and high-income group. Finally, disease type may affect the incidence of CHE.

The incidence of CHE in rural areas was significantly higher than that in urban areas. The income gap between urban and rural areas makes the proportion of rural households’ cash health expenditure higher than that of urban households, and the rural population faces greater CHE risks. Although the establishment and expansion of the New Cooperative Medical Scheme (NCMS) reduces the medical burden of poor rural residents [65], the NCMS is lower than the Urban Employee Basic Medical.

Insurance (UEBMI) in terms of fund-raising, security level and reimbursement scope. The combination of the two causes a higher risk of CHE in rural households. In addition, from 2009 to 2014, the average cost of inpatients nationwide increased from 5951.80¥ to 8290.50¥, an average annual increase of 6.86% [11]. The excessively rapid increase in medical costs partially offset the protective effect of the basic medical insurance system. The burden continues to increase.

The incidence of CHE in central areas was higher than that in eastern and western areas. The results of Li′s research found that the incidence of CHE is in the opposite direction to the regional economic level: that is, the incidence of CHE in the more economically developed areas is lower, the incidence rate in the eastern areas is the lowest, the central region is the second, and the western region is the highest [11]. This study is contrary to its results. The reason may be related to thecharacteristics of the research object, the representativeness of different literature samples is different.

The poverty group had the highest incidence of CHE, followed by the high-income group, followed by the low-income group, and the middle-income group had the lowest incidence of CHE. There are many reasons for this phenomenon. On the one hand, it may be passively over-utilized health services due to induced demand [66], that is,these families are more likely to accept more expensive drugs or over-checks provided by doctors, thereby increasing medical expenses [67]. This is a relatively common phenomenon in China; on the other hand, high-income groups may actively over-utilize health services [67], such as using higher-level nursing services or wards. This will also increase medical expenses and make families incur excessive cash health expenditures, thereby widening the overall disparity in CHE of middle-income groups. Of course, in the less developed areas of China, it may not be common for farmers to choose such special needs, but this cannot be ruled out. The poverty group will also be affected by induced demand, but due to economic reasons, its affordability for medical expenses (including the cost of induced demand) is limited.

When medical expenses exceed their ability to pay, poor families generally choose Abandon treatment [68]. It is also because of its low income that it cannot pay large amounts of health expenses, and the gap in catastrophic health expenditure may be reduced, but this also makes the use of normal health services for some poor people suppressed and its health level will be seriously affected.

Disease type may affect the incidence of CHE. Wang’s research shows that the risk of catastrophic health expenditures in families with chronically ill patients is generally higher than the average of the overall population, and there are inequalities based on income gaps [69].Wu’s research shows that there are inpatients at home, and the risk of catastrophic health expenditure is 2.5 times higher than that of the general population [70].

The overall incidence of CHE in the past 14 years (2006 to 2020) among Chinese residents was 23.3% (95% CI: 21.1 to 25.6%).Compared with other countries, the global incidence of CHE at the 10% threshold was estimated as 9.7% in 2000, 11.4% in 2005, and 11.7% in 2010. In 2010, the incidence of CHE in Asia was 3.1% [71]. In 2015, the incidence of CHE in South Korea was 2.4% [72], and in 2021, Iran’s survey results show that the incidence of CHE is 4.7% [73]. The possible reason for this situation is that this study did not distinguish patients from the general population, so it may overestimate the incidence of CHE.

Some specific measures targeting economic vulnerable groups are needed in order to reduce the incidence of CHE in China. First, policies are needed to increase the protection effects of the three types of social medical insurance, including scope and reimbursement ratio. While expanding the coverage of medical insurance in China, it needs deepen the coverage of medical insurance, such as increasing the actual compensation ratio of medical insurance;Expand the scope of medical insurance reimbursement and allow more expensive special-effect medicines and treatments to enter the medical insurance reimbursement catalog. Second, increase the scope and intensity of medical assistance and social assistance. Further strengthen the protection of the disadvantaged groups, chronic patients and other vulnerable groups, and give inclined support to the system. Establish medical assistance for severely and seriously ill patients to prevent them from becoming poor due to illness or becoming poor due to illness. Third, control the excessive growth of medical expenses.

Conclusion

Ten years after the new health reform in China, economic vulnerable groups still have higher risk of CHE than other groups. More research is needed to identify these population groups. And future health policies are warranted to contain healthcare cost, strength health financing protection and reduce the CHE in China towards universal health coverage.