Spatiotemporal characteristics of health expenditure in mainland China during 2012–2020

Abstract

Background

This study aims to examine the spatiotemporal characteristics of China’s total health expenditure structure from 2012 to 2020, and to observe the healthcare economic burden on the population in urban and rural areas across different provinces.

Methods

The data were obtained from the China Health Statistics Yearbook, the China Statistical Yearbook, and publicly available statistical data from various provinces, autonomous regions, and municipalities from 2012 to 2020. Joinpoint regression was employed to investigate the temporal trends of the three financing sources of total health expenditure (THE) and the total expenditure on health as percentage of gross domestic product (GDP). The elasticity coefficient of health consumption was calculated to measure whether there is a correlation between the growth of the health sector and that of the national economy. Furthermore, the global and local Moran’s I indices were used to examine the spatial characteristics of THE and its three financing sources in China from 2012 to 2020. A descriptive analysis was conducted to explore the characteristics of the healthcare economic burden in China and its provinces.

Results

From 2012 to 2022, China’s THE increased from 2,811.9 billion yuan to 7,217.5 billion yuan. The total expenditure on health as percentage of GDP exhibited a consistent upward trajectory, rising from 5.41% in 2012 to 7.12% by 2020. Over the past nine years, there has been a significant positive spatial autocorrelation in the total expenditure on health as a percentage of GDP, government health expenditure, and social health expenditure (Moran’s I > 0, P < 0.05). The Local Indicators of Spatial Association (LISA) cluster maps differ across these indicators. A consistent pattern emerged whereby urban populations spent significantly more annually on healthcare than rural populations. However, by 2020, this disparity had narrowed to 34.72%. Per capita medical care expenditure as a share of living expenditure of rural residents increased annually, reaching a threshold of 10% in 18 provinces by 2020. In contrast, urban areas remained below the 8% threshold.

Conclusion

Since 2012, China has consistently deepened its healthcare system reform, markedly increasing government investment in healthcare, particularly in impoverished areas, with the objective of achieving universal health coverage. Nevertheless, the economic burden of healthcare remains considerable for rural and underdeveloped populations. Furthermore, due to the dual factors of population aging and economic development levels, China’s THE will continue to face substantial financial pressure.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12913-026-14081-5.

Introduction

The growth trends of total health expenditure (THE), its financing characteristics, and the rationality and competitiveness of its structure profoundly influence the development of the healthcare sector [1]. These factors are critical for planning and allocating the resources needed to achieve universal health coverage and other health-related goals. THE is categorized by source into three components: government health expenditure (GHE), social health expenditure (SHE), and out-of-pocket health expenditure (OOP). According to the China Health Statistics Yearbook, THE represents the total monetary resources used for health services within a given period. GHE refers to government spending on healthcare services, health insurance subsidies, and public health administration; SHE includes social health insurance payments, employer and institutional contributions, and social donations; OOP refers to direct payments made by individuals when receiving healthcare services. This classification provides a macro-level reflection of the health financing levels and primary funding channels within a region over a specific period. It also illustrates the extent of commitment to healthcare by the government, society, and individuals under certain economic conditions, while highlighting the key characteristics and equity of the health financing model [2]. Analyzing the temporal and spatial trends of THE can help uncover regional and temporal disparities in healthcare financing and expenditure in China, providing a scientific basis for optimizing resource allocation and formulating more targeted health policies.

Economic development has significantly driven the growth of health expenditure [3]. Globally, health expenditure in both developed and developing countries has increased substantially over the past two decades [4–6], accounting for more than 10% of the world’s gross domestic product (GDP) [3], with the growth in developing countries being particularly pronounced [7]. Following China’s economic reforms and opening up, the rapid increase in THE has accompanied the rise in economic development and household incomes [8]. This has led to an increased demand for health services and higher expectations for the quality of healthcare. In 2018, THE in China reached 8.6% of GDP, with growth outpacing that of GDP, surpassing 8 trillion yuan in 2022 [9]. The government’s role in healthcare financing has steadily strengthened, with the proportion of government health expenditure in fiscal spending rising from 4% in 2007 to 7.8% in 2021 [10], reaching a total of 2.07 trillion yuan. Meanwhile, per capita health expenditure has increased to 6,010 yuan and is projected to grow to 14,129 yuan by 2030 [11].

Despite the remarkable achievements in healthcare spending within a relatively short period, challenges remain in alleviating the economic burden of healthcare across regions and between urban and rural areas, and in fully achieving the goal of universal health coverage [12]. Previous studies have shown a negative correlation between China’s economic development level and the economic burden of healthcare. The southeastern provinces bear a lower healthcare burden than other regions, while the central and western regions have higher rates of forgoing medical care due to economic reasons [13]. High-quality healthcare resources are predominantly concentrated in urban areas [14, 15], with healthcare financing and service utilization favoring urban residents [16]. In contrast, rural residents face an economic burden from healthcare expenditure that is 2.4 times greater than that of their urban counterparts, indicating persistent inequalities in healthcare spending [17].

Previous studies have provided valuable evidence on the evolution, spatial characteristics, and equity of China’s health expenditure, though they have predominantly focused on specific financing components or efficiency metrics. Zhang et al. utilized a Spatial Durbin Model to analyze OOP from 2005 to 2016, revealing that regional clustering is driven by demand-side factors and uneven resource distribution [18]. Wang et al. examined the efficiency of GHE between 2007 and 2016, finding significant spatial spillover effects where economic growth in one region positively influences the expenditure efficiency of its neighbors [19]. Additionally, Dong et al. analyzed the spatial-temporal differences in the OOP share of total health expenditure from 2013 to 2022, confirming that demographic structures and government spending continue to dictate provincial disparities in financial burden [20]. These studies have advanced understanding of China’s health expenditure, yet most relied on data ending before 2017 or focused on individual components such as GHE or OOP. This study extends previous work by analyzing the temporal trends and spatial patterns of THE and its structure, including GHE, SHE, and OOP, across 31 provinces in mainland China from 2012 to 2020. The findings reveal regional equity in health expenditure by comparing differences among provinces, providing updated evidence for improving health financing reform and optimizing resource allocation.

Methods

Data resource and design

The primary data of this study were obtained from the “China Health Statistics Yearbook 2022” [21]. Five indicators were selected to evaluate the overall trends of health expenditure in mainland China from 2012 to 2020, covering 31 provinces, autonomous regions, and municipalities in mainland China. Hong Kong SAR, Macau SAR, and Taiwan Province were excluded because data for these regions are not reported in the yearbook. (1) Total Health Expenditures (THE) (billions of Yuan), categorized into Government Health Expenditures (GHE), Social Health Expenditures (SHE), and Individual Cash Expenditure on Health (or out-of-pocket health expenditures, OOP), and their percentage compositions; (2) Total Expenditure on Health as Percentage of GDP (%); (3) Per Capita Expenditure on Health (yuan); (4) Urban residents’ health care expenditure as a percentage of consumer spending; (5) Rural residents’ health care expenditures as a percentage of consumer expenditures. Due to the unavailability of certain provincial-level data on total health expenditure financing (including the structure of government, social, and out-of-pocket expenditures) before 2012 and after 2020, the study period was restricted to 2012–2020 to ensure the accuracy and robustness of the analysis. Additionally, we collected Gross Domestic Product (GDP) data from the China Statistical Yearbook.

Statistical analyses

Annual percent change

Joinpoint 5.2.0 software was used to calculate the annual percent change (APC) and average annual percent change (AAPC) of China’s THE and its three financing sources from 2012 to 2020. This analysis aimed to assess overall and segmented time trends.

The elasticity coefficient of health consumption

The elasticity coefficient of health consumption is an important indicator used by countries worldwide to measure whether the development of healthcare is in harmony with economic growth. Generally, a health consumption elasticity coefficient slightly greater than 1 is required to ensure the steady development of the healthcare sector [22]. We used the 2022 “China Health Statistics Yearbook” and the 2023 “China Statistical Yearbook” to collect data on China’s THE and GDP from 2012 to 2020. The elasticity coefficient of health consumption was calculated using the following formula:

where ΔE represents the change in THE, E is the initial THE, ΔY represents the change in GDP, and Y is the initial GDP.

Global and local spatial autocorrelation tests

To identify geographic clusters of provinces with low or high prevalence, Moran’s test for global and local spatial autocorrelation analyses was employed. The global Moran’s I statistic [23] measures the similarity of attribute values among neighboring spatial units across the entire study area. The closer the absolute value is to 1, the stronger the relationship between the attributes of the spatial units. We used the global Moran’s I index to describe the spatial correlation of five provincial-level indicators of THE in China from 2012 to 2020. The global Moran’s I statistic for spatial autocorrelation strength is represented as follows:

where y_i and y_j represent the health expenditure in Province i and Province j, n is the total number of provinces, is the mean of observations, is the variance of the observation variable, and is the element in the spatial weights’ matrix. The Moran’s I value lies in the range of [-1,1]. A Moran’s I closer to 1 indicates a stronger positive spatial correlation between regions, whereas a value closer to -1 suggests a stronger negative spatial correlation. When Moran’s I equals zero, it indicates no spatial correlation among the observed values.

The local indicators of spatial associations (LISA) [24] are local indicators used to describe spatial relationships, where a positive value indicates a positive correlation between the attribute values of a spatial unit and its neighboring units, while a negative value indicates a negative correlation. We used LISA to test whether there was spatial autocorrelation in THE and its three financing sources among neighboring provinces in China from 2012 to 2020.

The LISA statistic for spatial autocorrelation strength is represented as follows:

where y_i, y_j, and are consistent with those used in the calculation of the Moran’s I index. A positive LISA value indicates similar spatial clustering patterns among adjacent regions, whereas a negative LISA value indicates differing patterns among these regions.

We identified provinces with high-high clusters, low-low clusters, high-low clusters, and low-high clusters. High-high and low-low clustering types indicate a strong positive correlation and spatial clustering in health expenditure levels. In contrast, high-low and low-high clustering types suggest a strong negative correlation and spatial differentiation.

Statistical analysis was performed using Stata MP 18.0 (Stata Corp., College Station, Texas, USA) software. The significance level for statistical tests was 0.05.

Results

Time trends in China’s health expenditure from 2012 to 2020

As shown in Fig. 1, THE in China rose from 2,811.9 billion yuan in 2012 to 7,217.5 billion yuan in 2020. Among these, SHE had the largest share, increasing from 35.70% to 41.94%, although it exhibited an overall decreasing trend (AAPC=-0.729). OOP increased from 965.63 billion yuan in 2012 to 1,995.94 billion yuan in 2020, but its share decreased from 34.30% to 27.65% (AAPC = -2.871***, 95% CI: -3.544, -2.050). The decline before 2015 (APC=-5.501***) was significantly faster than after 2015 (APC=-1.258). GHE increased from 843.2 billion yuan in 2012 to 2,194.2 billion yuan in 2020. Its share decreased from 30% in 2012 to 27.36% in 2019, but due to the COVID-19 pandemic in 2020, it rose again to 30.4% (Fig. 1 and Supplementary Table 1). From 2012 to 2019, the elasticity coefficient of health consumption of China fluctuated between 1.18 and 2.28. In 2020, the impact of the COVID-19 pandemic caused a significant slowdown in GDP growth, while expenditures related to pandemic control increased sharply, driving a high rate of growth in health expenditure. Consequently, the health consumption elasticity coefficient for that year reached 3.51. (Supplementary Table 2)

Fig. 1.

Total health expenditure and its three financing sources in China from 2012 to 2020

Percentage of the and its three financing sources to GDP from 2012 to 2020

The total expenditure on health as percentage of GDP in China has been steadily increasing, rising from 5.41% in 2012 to 7.12% in 2020. At the provincial level, in 2012, the total expenditure on health as percentage of GDP showed a higher trend in the west compared to the east. For instance, the THE in Xizang(Tibet), Gansu, and Xinjiang accounted for 9.13%, 7.87%, and 7.55% of their respective GDPs, while in Fujian, Jiangsu, and Tianjin, it accounted for 3.44%, 3.50%, and 3.72%, respectively. In Beijing, THE was 119 billion yuan, accounting for 6.66% of GDP, ranking 25th. Zhejiang had the lowest percentage of GHE at 22.2%, while Xizang had the highest at 68%. Meanwhile, Xizang had the lowest percentage of OOP at 7.4%, with per capita expenditure on health at 2,079.51 yuan.

In 2020, Guangxi had the lowest per capita expenditure on health (3,739.2 yuan), while Beijing had the highest (13,834.01 yuan). The total expenditure on health as percentage of GDP in Fujian and Jiangsu was less than 5%, whereas in Xinjiang, Xizang, Gansu, Qinghai, and Heilongjiang, it exceeded 10%. Xizang had the highest percentage of GHE (69.91%) and the lowest percentage of OOP (7.16%) and SHE (22.93%). Guangdong had the highest SHE, reaching 329.36 billion yuan, while Beijing had the highest percentage of SHE at 59.86%, an increase of 9.36% from 2012 (Fig. 2).

Fig. 2.

The total expenditure on health as percentage of GDP, percentage of GHE and OOP in various provinces of China in 2012 and 2020. (A) Percentage of Total Health Expenditure to GDP in Various Provinces of China in 2012 (%); (B) Percentage of Government Health Expenditure in Various Provinces of China in 2012 (%); (C) Percentage of Out-of-pocket Health Expenditure in Various Provinces of China in 2012 (%); (D) Proportion of Total Health Expenditure to GDP in Various Provinces of China in 2020 (%); (E) Percentage of Government Health Expenditure in Various Provinces of China in 2020 (%); (F) Percentage of Out-of-pocket Health Expenditure in Various Provinces of China in 2020 (%)

Global and local spatial autocorrelation

After calculating the Global Moran’s I, we found significant positive spatial autocorrelation in the total expenditure on health as percentage of GDP, the percentage of GHE, and the percentage of SHE from 2012 to 2020 (Moran’s I > 0, P < 0.05). Per capita health expenditure also exhibited significant positive spatial autocorrelation from 2012 to 2019 (Moran’s I > 0, P < 0.05), while the percentage of OOP showed significant positive spatial autocorrelation from 2012 to 2018 (Moran’s I > 0, P < 0.05) (Table 1). Among these, the total expenditure on health as percentage of GDP had the strongest spatial correlation from 2012 to 2020. The spatial correlation of the percentage of OOP ranked second from 2012 to 2015, but starting from 2016, the spatial correlation of the percentage of SHE and GHE surpassed that of OOP.

Table 1.

Global Spatial autocorrelation results of total health expenditure and various indicators in China (2012–2020)

Indicators	2012	2013	2014	2015	2016	2017	2018	2019	2020
The total expenditure on health as percentage of G DP	0.429***	0.465***	0.402***	0.491***	0.537***	0.587***	0.526***	0.502***	0.469***
Per capita expenditure on health	0.259**	0.262**	0.272**	0.259**	0.228**	0.201*	0.169*	0.153*	0.119
The percentage of government health expenditure	0.270**	0.242*	0.272**	0.255**	0.218*	0.213*	0.224*	0.225*	0.354***
The percentage of social health expenditure	0.291**	0.260*	0.257*	0.257*	0.252*	0.260*	0.265*	0.288**	0.415***
The percentage of out-of-pocket health expenditure	0.320**	0.266**	0.323**	0.294**	0.207*	0.201*	0.182*	0.166	0.114

***P < 0.001, **P < 0.01, *P < 0.05

Further analysis using the local Moran’s I index to characterize observations between regions showed that, in 2020, the “high-high” clusters for the total expenditure on health as percentage of GDP were found in Xizang, Gansu, Qinghai, and Xinjiang, all located in the western region of China, while the “low-low” clusters were in Jiangsu, Zhejiang, and Fujian. Spatial autocorrelation in other regions was not significant (Fig. 3A, Supplemental Fig. 1). Tianjin was identified as a “high-high” cluster for per capita health expenditure (Fig. 3B, Supplemental Fig. 2). The “high-high” clusters for the percentage of GHE were located in Yunnan, Xizang, Qinghai, and Xinjiang, with Jiangsu being a “low-low” cluster (Fig. 3C, Supplemental Fig. 3). In terms of the percentage of SHE, Beijing, Tianjin, Shanghai, and Jiangsu formed “high-high” clusters, while Yunnan, Xizang, Qinghai, and Xinjiang were “low-low” clusters (Fig. 3D, Supplemental Fig. 4). For the percentage of OOP, Beijing was a “low-high” cluster, and spatial autocorrelation in other regions was not significant (Fig. 3E, Supplemental Fig. 5).

Fig. 3.

Regional Spatial Autocorrelation Clusters of Various Health Expenditure Indicators in China, 2020. (A) Regional Spatial Autocorrelation Clusters of the Total Expenditure on Health as Percentage of GDP in 2020; (B) Regional Spatial Autocorrelation Clusters of per capita Expenditure on Health in 2020; (C) Regional Spatial Autocorrelation Clusters of the Percentage of Government Health Expenditure in 2020; (D) Regional Spatial Autocorrelation Clusters of the Percentage of Social Health Expenditure in 2020; (E) Regional Spatial Autocorrelation Clusters of the Percentage of Out-of-pocket Health Expenditure in 2020

Percentage of household healthcare expenditure to total consumption expenditure in urban and rural areas in China from 2012 to 2020

Generally, annual THE for urban populations is significantly higher than that for rural populations (Supplementary Fig. 6), but this gap narrowed to 34.72% in 2020. The percentage of healthcare expenditure to total consumption expenditure for rural residents has increased annually, surpassing 10% in 2018, while for urban residents, this percentage has remained below 8%. At the provincial level, in 2012, the percentage of healthcare expenditure to total consumption expenditure for urban residents in all provinces of China was below 10%. However, by 2020, this percentage exceeded 10% in Shanxi (11.9%), Liaoning (10.4%), Jilin (11.1%), Heilongjiang (11.5%), Shaanxi (11.4%), Qinghai (10.4%), Ningxia (10.1%), and Xinjiang (10.2%). For rural areas, only three provinces had a percentage below 10% in 2012, but this number increased to 19 in 2019 and slightly decreased to 18 in 2020. Liaoning had the highest percentage at 14%, followed by Shaanxi at 13.1% (Fig. 4).

Fig. 4.

Percentage of household healthcare expenditure to total consumption expenditure for urban and rural populations in various provinces of China from 2012 to 2020

Discussion

This study, based on China’s health account data, describes the spatiotemporal trends in health expenditure from 2012 to 2020. We incorporated the elasticity coefficient of health consumption and spatial autocorrelation methods to explore the coordination between expenditure growth and economic development as well as regional clustering patterns. Our findings indicate that China’s THE has increased annually, with its share of GDP also rising. However, greater attention needs to be paid to health expenditure in less developed regions and rural areas.

In terms of temporal trends, THE in China steadily increased from 2012 to 2020, reaching 7.22 trillion yuan in 2020, with its percentage of GDP rising from 5.41% to 7.12%. This suggests that the overall level of health investment in China has gradually improved. Some studies predict that the total expenditure on health as percentage of GDP will continue to expand in emerging economies, including China [25]. Unlike OECD countries where health spending growth has largely stabilized, China’s rapid expansion mirrors the trajectory of other transitional economies but at a more accelerated pace due to active government intervention and the rapid expansion of social health insurance coverage [26]. Aside from the impact of the COVID-19 pandemic in 2020, the elasticity coefficient of health consumption in China fluctuated between 1.18 and 2.28 from 2012 to 2019, indicating that the growth rate of health expenditure far exceeded the economic growth rate during this period. This suggests that the health financing system may face increasing pressure under the evolving economic structure and constrained fiscal capacity in China [27]. Studies have shown that the increase in health expenditure is primarily due to rising incomes [28, 29] and technological advancements [30, 31]. Economic growth raises household income levels, enabling a country to allocate more resources to healthcare services [32]. Additionally, advances in medical technology and the comprehensive coverage of health insurance have further promoted the extensive use of healthcare services, driving up THE [29]. However, a previous research suggests that increasing health expenditure alone is not a representative indicator of improved health outcomes [33]. For example, in BRICS countries, the rise in health expenditure is primarily driven by population aging and the rapid increase in non-communicable diseases (NCDs) [16, 34]. Therefore, it is important to view the rise in THE with caution and to develop health investment policies and measures that are tailored to local conditions. Research has shown that China’s public health system and health insurance system suffer from significant resource wastage, with induced examinations and high OOP expenses placing financial pressure on both patients and the insurance system [35, 36]. Therefore, rather than blindly pursuing increased health expenditure, more emphasis should be placed on the efficient use of health funds. Measures should be taken to address the key factors contributing to the unreasonable growth of costs, optimize resource allocation, and achieve more efficient and equitable healthcare services. It should also be acknowledged that the observed expenditure growth does not necessarily reflect actual healthcare needs or service accessibility, especially given the persistent disparities between urban and rural areas.

When THE was categorized by source, we found that the percentage of OOP decreased annually, while the percentage of SHE increased. This aligns with the health financing transition theory proposed by Fan and Savedoff [37] and is consistent with previous findings on China’s gradual shift toward lower OOP and higher social health financing [20, 38]. The percentage of OOP is considered an effective indicator of the economic burden on individuals and the equity of health financing [39]. Over the past nine years, the percentage of OOP in China decreased from 34.30% to 27.65%, achieving the “around 28% by 2020” target set in the “Healthy China 2030 Plan” [40], indicating an improvement in risk resistance capacity. This achievement can be attributed to the universal coverage of health insurance, the enhanced capacity for health insurance financing, and the continuous improvement of the national essential drug system [41]. However, a decline in out-of-pocket payments does not necessarily imply a reduction in the overall economic burden of healthcare. Variations in service prices, reimbursement levels, and benefit coverage across regions may continue to impose financial stress on households. A recent study has shown that although the share of OOP has gradually decreased nationwide, interprovincial disparities persist, with economically developed regions benefiting more from health system reforms, while residents in less developed areas continue to face relatively higher financial pressure for healthcare [20]. Moreover, although financial protection has improved at the aggregate level, households in less developed regions may still face a higher risk of catastrophic health expenditure [42].

In terms of spatial distribution, both the total and structural indicators of THE showed significant spatial clustering, indicating a clear regional distribution pattern in China’s health expenditure. This positive spatial autocorrelation suggests that provinces with high levels of health expenditure tend to be geographically adjacent to other high-expenditure provinces, while provinces with lower expenditure are clustered together, reflecting spatial convergence among regions with similar economic development and fiscal capacity. The differences in health investment across regions are somewhat inevitable due to the varying stages of economic and social development, disease profiles, aging populations, and the pace of urbanization in different regions of China [11]. However, effective measures are still needed to bridge the gaps in health equity. Research has shown that China’s healthcare reforms have made initial progress in promoting health equity, with the clustering of per capita healthcare connectivity gradually easing. However, resources remain concentrated in wealthier regions, further widening the gap in healthcare utilization between the rich and the poor [43]. At the provincial level, our study found that the total expenditure on health as percentage of GDP consistently showed a “west-high, east-low” pattern, with varying levels of GHE across different regions, consistent with previous research [18]. Some provinces in the western region, such as Xinjiang, Qinghai, Xizang, and Yunnan, have slower development due to geographic constraints. Although the percentage of SHE is low in these regions, the government has provided substantial financial support, allocating sufficient special funds through fiscal transfer payment policies, thereby reducing the percentage of OOP and significantly improving health equity [44]. In contrast, economically developed regions in the eastern and coastal areas, such as Beijing, Tianjin, Shanghai, and Jiangsu, have relatively low percentages of GHE. However, these regions have well-established social health insurance systems, strong social insurance financing capacity, and high percentage of SHE, resulting in lower OOP. These regional variations in expenditure structure may also be influenced by differences in healthcare service availability, local price levels, and the technological composition of medical services [26]. For example, lower OOP shares in eastern provinces may reflect better insurance coverage and more efficient purchasing mechanisms rather than lower healthcare costs [18]. However, we found that in 2020, 14 provinces had a percentage of OOP exceeding 28% of THE, and in seven less developed provinces (Shanxi, Liaoning, Jilin, Heilongjiang, Shaanxi, Qinghai, and Ningxia), the average OOP exceeded 10% of household annual living expenses. This indicates an urgent need to strengthen health funding in less developed regions, and that the affordability and equity of healthcare require further improvement [2]. Therefore, the central government should enhance top-level design, increase fiscal transfer payments to less developed regions [45], and raise the level of GHE while providing legal guarantees for GHE, clarifying the responsibility of investment at all levels of government, and implementing a supervision system. Each region should adopt a comprehensive approach to further enhance risk protection capacity in financing. In particular, provinces in the central and western regions need to maintain the existing growth rate of GHE, strengthen interprovincial exchanges and cooperation, and draw on successful examples from developed cities and coastal provinces with well-established social health insurance systems. They should also work to further improve the multi-level healthcare system, promote the development of commercial health insurance, refine the third distribution system for social charity donations, and explore new sources of financing to increase the percentage of SHE in THE, thereby reducing the percentage of OOP [46].

Our study found that although the gap in per capita healthcare expenditure between urban and rural residents is narrowing, the gap in the percentage of per capita expenditure on health to per capita disposable income is widening. In 2020, the average OOP exceeded 10% of household annual living expenses in 18 provinces. This may be related to inequities in access to healthcare between urban and rural residents, with urban residents benefiting more from the Urban and Rural Resident Basic Medical Insurance (URRBMI) system. URRBMI was launched nationwide in 2016 by integrating the Urban Resident Basic Medical Insurance (URBMI) and the New Cooperative Medical Scheme (NCMS) for rural populations, aiming to improve financial protection and reduce inequities in healthcare utilization between rural and urban populations [47]. However, differences in financing capacity and local implementation have led to persistent disparities in reimbursement levels and benefit coverage. This observation is consistent with recent study, which suggest that although the reform improved benefit levels and expanded reimbursement coverage, its effect on reducing regional and income-related disparities in health and financial protection remains limited [48]. Future efforts should focus on adjusting the financing policies of URRBMI to address the issue of “reverse subsidies” [49]. Research has shown that in the context of aging, the healthcare needs of urban and rural residents differ. Urban aging has a positive impact on per capita expenditure on health, while the healthcare needs of the rural elderly population have not been adequately met [11]. These findings suggest that regional variations in actual healthcare spending may also indicate different levels of realized healthcare demand, which are influenced by disparities in local economic capacity and healthcare service accessibility. Previous studies have also reported that rural households in central China are more likely to experience catastrophic healthcare costs than urban households in eastern China [50]. This suggests that more attention should be paid to the equity of health financing in less developed regions and rural areas. The central and local governments should take the initiative to innovate within the system, manage limited resources, explore new financing models, and improve the resilience and efficiency of healthcare systems in resource-scarce areas. This would accelerate the implementation of high-quality universal health coverage (UHC) and mitigate the significant risk of impoverishment due to illness [51].

This study systematically analyzed the temporal trends and spatial characteristics of health expenditure in China from 2012 to 2020, providing valuable guidance for optimizing China’s health financing structure and informing GHE decisions. In the future, China needs to promote coordinated development between the economy, society, and the healthcare sector, continue deepening healthcare reform, enhance government initiative and innovation, and leverage its role in basic health services and health equity. The government should use social redistribution mechanisms to reduce urban-rural and regional disparities and increase investment in health expenditure in less developed regions and rural areas, while also improving monitoring of the efficiency of this investment. Additionally, efforts should be made to improve social health investment policies, further strengthen the universal health insurance system, increase the pooling level of basic health insurance, reduce regional policy fragmentation in social healthcare programs that hampers portability and equity of coverage, promote the balanced development of commercial health insurance across regions, and expand government and social health financing channels. The goal is to establish a diversified, efficient, and sustainable financing mechanism, gradually correcting imbalances in resource allocation and improving the equity of health investment across regions. Meanwhile, actuarial studies of funds should be conducted, and monitoring and early warning systems for THE should be strengthened. Efforts should also be made to enhance the performance evaluation of health investment, establish a standardized and transparent fund management system, and improve the efficiency of fund utilization.

This study, based on national health statistics, describes the overall trends in GHE, SHE, and OOP as well as the trends in regions with different levels of socio-economic development. However, this study has certain limitations. First, the national THE data are derived from aggregate population data, making it difficult to understand the specific population distribution and their healthcare-seeking behaviours, which limits the study’s ability to accurately reflect individual-level health equity issues. Second, due to differences in the years when different regions of China began accounting for THE, we were unable to obtain complete data for 31 provinces in China mainland before 2012 and after 2020. Additionally, the special events of 2020 may have led to an overestimation of the annual growth rate of health expenditure. Moreover, this study was unable to obtain information on potential confounding factors, making this an unadjusted analysis. The limited range of socio-economic factors included in this study may influence other factors affecting health expenditure (such as healthcare infrastructure, transportation, and housing), which require further research. Finally, this study focuses on the financing and structural characteristics of health expenditure and therefore does not examine the association between spending patterns and population health outcomes. Future research could integrate health outcome indicators to explore how variations in expenditure structure relate to population health improvements.

Conclusion

Since 2012, China has continuously deepened reforms in healthcare system, significantly increasing government investment in health, particularly in impoverished areas, to achieve UHC. However, attention should continue to be focused on the equity of healthcare financing for rural populations, especially those in underdeveloped regions. In the future, efforts should continue to be guided by the health needs of the population. It is essential to rationally adjust the percentages of GHE, SHE, and OOP, accelerate the construction of a diversified healthcare financing mechanism, strengthen monitoring and evaluation of health investment performance, enhance healthcare system efficiency, and increase health expenditures sensibly. There should be a focus on enhancing investments in remote and impoverished areas, mobilizing local governments’ enthusiasm in healthcare system reforms, promoting healthy development of the health financing structure, and more efficiently utilizing healthcare financial resources. These measures will further improve fairness and accessibility of medical services.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

AAPC

Average Annual Percent Change

APC

Annual Percent Change

BRICS

Brazil, Russia, India, China, and South Africa

CI

Confidence Interval

GDP

Gross Domestic Product

GHE

Government Health Expenditure

LISA

Local Indicators of Spatial Association

NCDs

Non-Communicable Diseases

NCMS

New Cooperative Medical Scheme

OECD

Organisation for Economic Co-operation and Development

OOP

Out-of-Pocket Health Expenditure

SAR

Special Administrative Region

SHE

Social Health Expenditure

THE

Total Health Expenditure

UHC

Universal Health Coverage

URBMI

Urban Resident Basic Medical Insurance

URRBMI

Urban and Rural Resident Basic Medical Insurance

Author contributions

All authors helped develop the study concept and design. YB, QW, QL, SL, YF, JS and MW contributed to the data collection, statistical analysis and manuscript writing. MW, YJ and MR provided overall guidance and critical revision. All authors revised the manuscript and approved the final version. All authors declare that none of this manuscript was written using AI.

Funding

This study was funded by the National Natural Science Foundation of China (No. 72274005) and Beijing Nova Program (20230484284). The study sponsor had no role in the study design, data analysis and interpretation of data, the writing of the manuscript, or the decision to submit the paper for publication.

Data availability

The primary data of this study were obtained from the “China Health Statistics Yearbook 2022”. Additionally, we collected Gross Domestic Product (GDP) data from the China Statistical Yearbook. All data are publicly available on CNKI Big Data Research Platform.

Declarations

Ethics approval and consent to participate

Ethics approval and consent to participate were not required for this study.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.