Efficiency and equity of hospital services in mainland China: based on data from 2017 to 2021

doi:10.21203/rs.3.rs-2590008/v1

Download PDF

Research Article

Efficiency and equity of hospital services in mainland China: based on data from 2017 to 2021

https://doi.org/10.21203/rs.3.rs-2590008/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background: Hospitals are an essential part of the health care system. The purpose of this study is to evaluate the equity and service efficiency of health resource allocation in the hospital service system of 31 provinces (municipalities directly under the central government and autonomous regions) in mainland China from 2017 to 2021, and to analyze the influencing factors of hospital service efficiency to provide a reference for rational allocation of hospital health resources and improvement of hospital service efficiency.

Methods: Equity of allocation of hospital health resources in mainland China was evaluated in three dimensions using the Gini coefficient and Lorenz curve. Hospital service efficiency was calculated by output-oriented super efficiency slack-based model-data envelopment analysis (SE-SBM-DEA) model and DEA-based Malmquist total factor productivity index (MTFP), and the influencing factors of hospital service efficiency were analyzed by panel data regression.

Results: The number of hospitals in mainland China grew from 31056 to 36570 between 2017 and 21st. The equity ranking of the dimensions of hospital health resource allocation was Demographic > Economic > Geographic. However, the curvature of the demographic and economic dimensions is smaller compared to the curvature of the Lorenz curve for the geographic dimension. The demographic dimension has the lowest average Gini coefficient (0.167), followed by the economic (0.295) and the geographical (0.616). The 2017-2021 MTFP results show that none of the DMU efficiency values exceed 1. SE-SBM-DEA results show 15 DMUs in 2021 with total factor productivity above 1. The five-year average efficiency values ranked West (0.851) > East (0.834) > Central (0.673); Five influential factors identified by panel data regression analysis, Among them, X1(The proportion of urban population), X3(The percentage of illiterate population to total aged 15 and over) and X10(The proportion of the volume of medical service in primary medical facilities) have a positive relationship with hospital service efficiency, and X5(The level of per capital outpatient and emergency treatment expenditure) and X8(The proportion of hospital personnel) have a negative relationship with hospital service efficiency.

Conclusion: After five years of development, the hospital service system in mainland China has grown significantly in number and size, but there are still differences in the allocation of hospital health resources between the East and the West, and intra-regional differences remain large. The overall level of efficiency in hospital services over the last five years was poor, with the West overtaking the East as the region with the highest five-year average. The regression model verified the effect of some factors on hospital service efficiency, and the effect of disadvantaged groups on hospital service efficiency was not significant (P>0.05).

hospitals

allocation

equity

efficiency

influence factors.

Hospital spending accounts for about 30–80% of total health spending in all countries [1]. Hospitals continue to be placed at the center of cost containment policies by governments because of their large share and their apparent operational inefficiencies [2]. All countries in the world attach great importance to the study of health care resources, and how to allocate health care resources rationally and effectively and improve the efficiency of services has become the primary concern of health reform in the world [3–6]. Therefore, scientific evaluation of the efficiency of the hospital service helps to optimize the allocation of health resources and thus meet the health needs of the people [7].

Hospitals, as the backbone of the Chinese health system, are a major force in the implementation of China's health strategy. During the three-year period of COVID-19 epidemic prevention and control in China, the proportion of patients admitted to hospitals in all regions of the country exceeded 98%, reflecting the essential role of hospitals in safeguarding people's lives and health and social stability [8]. Health resources are the foundation for the provision of health services, and the rational allocation and efficient use of hospital health resources is a prerequisite for improving health output and health service levels. China, as a developing country, is at the same time a populous country. Health policy makers face daunting challenges in seeking to meet the health care needs of this large group [9].

The “New Medical Reform” introduced in 2009 clarified the public welfare of health care and emphasized the unified pursuit of equity and efficiency [10]. In 2021, the “Fourteenth Five-Year Plan Outline (2021–2025)" and “Vision to 2035” were introduced, emphasizing the need to improve the quality and efficiency of healthcare, expand the supply of healthcare service resources, and accelerate the expansion and balanced regional distribution of high-quality healthcare resources [11]. In particular, the "Health China 2030" plan proposed by the Chinese government advocates "accelerating the expansion and balanced regional distribution of high-quality health resources", which is regarded as a breakthrough for health improvement. Researchers usually use the Gini coefficient, the Thayer index, the concentration index, and the Lorenz curve to measure equity in the distribution of health resources [12].

The Lorenz curve was proposed in 1907 by Austrian statistician M.O. Lorentz. The Gini coefficient was introduced in 1912 by Italian economist Corrado Gini. The combination of the Lorenz curve and the Gini coefficient can vividly reflect the equity of resource allocation. It is now widely used in the study of resource allocation equity. For example, Zhang [13] used Lorenz curve, Gini coefficient, Theil index and health resource density index (HRDI) to assess the equity of primary health care resources in demographic and geographical dimensions. DEA and MTFP were chosen to measure equity and efficiency in primary health care resource allocation in mainland China; Toyabe [14] used the Gini coefficient, Atkinson index, and Theil index as indicators of unequal distribution of physicians in the population; Ding [15] conducted a comparative study of the equity and efficiency of provincial health care delivery systems in mainland China from 2009 to 2014 using the Gini coefficient and Lorenz curve, pointing out significant differences in the allocation of health resources between regions.

Data Envelopment Analysis (DEA) is now proving to be the method of choice for measuring performance and assessing low productivity in the hospital industry. DEA is a widely used nonparametric method whose main advantage is that it does not require assumptions about the form of the function and can include multiple inputs and outputs measured in different units at the same time [16–17]. Nunamaker (1983) [18] first applied DEA to the medical field, and in 1984, Sherman [19] first applied DEA to hospital level for efficiency evaluation. Meanwhile, several classical DEA models were widely used worldwide, including Charnes, Cooper, and Rhodes (CCR, 1978) [20], Banker, Charnes, and Cooper (BCC, 1982) [21], and later gradually derived from multi-stage DEA models, slack-based measure (SBM) models and super-efficiency SBM-DEA models to evaluate and analyze efficiency [15, 22–23]. As well as dynamic models such as Dt-SBM model and Malmquist index for inter-period efficiency evaluation and analysis [24–25]. The DEA family has gradually grown larger and more mature.

Over the last 30 years, there has been a wealth of research into the efficiency of hospital services. Hou [26] evaluated the static and dynamic efficiency of resource allocation in health care institutions by combining SE-SBM-DEA and MTFP; Mohamadi [27] evaluated and analyzed the performance of the Iranian health system on the efficiency of health resources use using the Malmquist index model; Giménez [28] calculated the directional distance function and the global Malmquist-Luenberger index to analyze the efficiency of hospitals in the Colombian health system from both static and dynamic perspectives; some scholars have also used multi-stage DEA models, the slack-based measure (SBM) model and SE-SBM-DEA to evaluate and analyze efficiency [15, 23]; They applied regression analysis to determine the impact of population, per capita income, number of beds and medical professionals, and total expenditures on DEA scores.

The aim of this study is to assess the equity and service efficiency of the allocation of health resources in the mainland Chinese hospital service system from 2017 to 2021, while identifying its impact factors. The study is divided into the following six sections: The first section begins with an introduction to the background of the paper and the current status of the research. The second is the source of the data, the introduction of the methodology, and the selection of indicators and variables. Section 3 is an analysis of the allocation of hospital health resources in China using the Gini coefficient and the Lorenz curve. The fourth section is devoted to the efficiency analysis of the collected input and output variables from static and dynamic perspectives using SE-SBM-DEA and MTFP, respectively. Section 5 is a panel regression analysis of the resulting values of the SE-SBM-DEA with panel data for each of the influencing factors to identify factors that may affect the efficiency of hospital services. Section 6 is devoted to the analysis and discussion of the results.

2.1 Data source

In this research, hospitals in mainland China from 2017 to 2021 were used as the study population, and data related to hospitals in 22 provinces, 5 autonomous regions, and 4 municipalities directly under the central government were collected (Taiwan, Hong Kong, and Macao were not included because of inconsistent statistical standards). The hospital service systems in this study were defined as general hospitals, TCM hospitals, combined TCM hospitals, ethnic hospitals, and various specialty hospitals, excluding specialty disease control hospitals, maternal and child health care hospitals, and nursing homes. The data required was obtained from the 2018–2022 Chinese Yearbook of Health and Health Statistics. The year-end total population data and regional GDP data for each region are from the China Statistical Yearbook 2018–2022.

The definition of the three major health regions in China was used in the Chinese Health and Health Statistics Yearbook to divide the health regions. Delineation criteria: the eastern part includes 11 regions (Beijing, Tianjin, Hebei, Liaoning, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong, Hainan); the central part includes 8 regions (Shanxi, Jilin, Heilongjiang, Anhui, Jiangxi, Henan, Hubei, Hunan); the western part includes 12 regions (Inner Mongolia, Chongqing, Guangxi, Sichuan, Guizhou, Yunnan, Tibet, Shaanxi, Gansu, Qinghai, Ningxia, Xinjiang). The subjects of this paper are hospital service systems in 31 regions of mainland China. The hospital service system can be seen as a production process with numerous different inputs and outputs, and it can be considered as a macro-healthy production unit [15]. Therefore, the hospital service system of each provincial unit in China can be considered as a decision-making unit (DMU).

2.2 Research Design

2.2.1 Stage 1: Equity of hospital health resource allocation

In the methodological section, it is divided into three subsections. The first stage is to describe the equity of resource allocation of hospital service system in mainland China using Gini coefficient and Lorenz curve, which are analyzed in three dimensions (economic, geographic, and demographic). A Lorenz curve is a curve used to reflect the fairness of resource allocation, mainly based on the arcs of the curve to reflect the level of fairness. The greater the arc of the straight line, the worse its level of equity [29]. In this paper, the curve formed by the number of hospitals in each region as a percentage of the total number of hospitals in the country and the corresponding population/economy/geographic as a percentage of the total population/economy/geographic [30].

Specifically, the line formed by the cumulative percentage of the population/economy/geography of each region as the horizontal coordinate and the cumulative percentage of the number of hospitals in each region as the vertical coordinate in a two-dimensional plane and its corresponding scatter are called Lorenz curves. The Gini coefficient is one of the metrics that quantitatively assess the equity of resource allocation. Mr. Zhang of Shanxi Agricultural University in China then summarized a simpler formula: the cumulative percentage of population/economy/geography in each region is divided into n groups by arranging the percentage composition of the study subjects from smallest to largest. Assuming that the cumulative percentage share of group 1 to group i is Yi, the Lorenz curve will pass through the point (i, Yi). Defining Y0 = 0 and Yn = 1 will give i = 0,1,2, ... n, for a total of n + 1 points on the Lorenz curve. Then use the trapezoidal rule to integrate the area under the Lorenz curve to find the area of the right triangle, and finally get the Gini coefficient as A/(A + B) = 1-2B. The Gini coefficient takes a value between 0 and 1, and the closer it is to 0, the better the fairness. In general, its value below 0.2 means the absolute mean, and values greater than 0.5 indicate high inequity [13]. In this paper, we use the Lorenz curve and the Gini coefficient to measure equity in the distribution of hospital health resources.

2.2.2 Stage 2: hospital service efficiency evaluation

The selection of metrics for the inputs and outputs of these regional hospitals is first presented in this section. In selecting input and output indicators for hospital service systems, we followed the examples of other studies involving health care systems [15, 31–32] and selected three inputs and three outputs to describe hospital service efficiency in 31 regions. Regarding the selection of input-output indicators, most studies selected the corresponding indicators from the three health resources: human, financial and physical [33–35]. The number of health care facilities and the number of beds were generally chosen to represent capital investment. Health workers, health technology workers, and physicians are used as human investments; for the selection of output indicators, most choose the number of outpatient visits or outpatient and emergency visits, and hospitalizations or discharges [3, 36–39]. In this study, based on the existing literature and the corresponding indicators available, the number of hospitals (I1), the number of health technicians (I2) and the actual number of beds (I3) were selected as input indicators, and the number of visits (number of outpatient and emergency treatment) (O1), discharges (O2) and bed utilization rate (O3) were selected as output indicators in two dimensions: service volume and economic efficiency (Table 1). According to the statistics of the Chinese Ministry of Health, health technicians in the text include doctors, nurses, other clinical staff, etc., excluding administrative staff and other non-clinical staff engaged in management.

The efficiency of the above input and output is then evaluated using SE-SBM-DEA and MTFP, respectively. Traditional models include CCR, BCC and other models, which have the following two defects: first, they are all measured based on a radial perspective, without considering the slack variables of the input and output components, and the disproportionate changes of the output quantities such as the number of consultations and beds in the hospital service efficiency lead to inaccurate measurement results. Second, the measured efficiency values range from 0 to 1. The DMUs that reach the efficiency value 1 simultaneously are not comparable to each other.

Tone K [40] proposed a non-radial, non-angle slack-based measure (SBM) model based on the traditional DEA model, which can effectively compensate for the shortcomings of some traditional models that do not incorporate slack variables when measuring low efficiency and improve the accuracy. Based on this, Tone K combined the SBM model with the super-efficiency model [41] and proposed the super efficiency slack-based model-data envelopment analysis (SE-SBM-DEA) model, which can further compare decision units (DMU) that simultaneously reach an efficiency value of 1 to compensate for the second deficiency mentioned above.

Therefore, in this study, the SE-SBM-DEA model based on input volume measurements is used to accurately reflect the efficiency of hospital services in 31 provinces of China. The formula is shown in (1).

n is the number of measured regions (n = 31 in this study), j is the j-th province (j = 1, 2, ..., n), and each region has m inputs and z outputs. ${x}_{ij}$and ${y}_{rj}$ denote the i-th input and r-th output of the j-th region. ${x}_{ik}$ and ${y}_{rk}$denote the i-th input and r-th output of the k-th region. ρ* denotes the efficiency value of each region. λ denotes the weight vector; ${s}_{i}^{-}$and ${s}_{i}^{+}$are the slack variables of the input vector and output vector, which denote the savings of hospital health resources inputs and the increase of hospital service outputs in the pilot regions, respectively. In addition, according to model (1), the hospital service efficiency in 31 provinces in China from 2017–2021 can be calculated. As shown in Eq. (2), ${Score}_{kt}$denotes hospital service efficiency in region k in year t.

$${Score}_{kt}={\rho }_{kt}^{\text{*}}\left(k=\text{1,2},3\dots ..31;t=\text{2017,2012},\dots .,2021\right)$$

In terms of efficiency evaluation, the SE-SBM-DEA model can only measure the efficiency within a certain period of time. It cannot consistently compare different time periods and does not reflect the dynamic evolution of efficiency. Therefore, this study uses the MTFP model to further calculate the total factor productivity (TFP) of Hospital Service services to reveal its dynamic evolution.

Let (${x}^{t}$,${y}^{t}$,) and (${x}^{t+1}$, ${y}^{t+1}$) denote the input-output vectors of the hospital service system in periods t and t+1, respectively, the TFPCH of the two adjacent periods are:

$$TFPCH=M\left({x}^{t},{y}^{t},{x}^{t+1},{y}^{t+1}\right)={\left[\frac{{D}^{t}\left({x}^{t+1},{y}^{t+1}\right)}{{D}^{t}\left({x}^{t},{y}^{t}\right)}\times \frac{{D}^{t+1}\left({x}^{t+1},{y}^{t+1}\right)}{{D}^{t+1}\left({x}^{t},{y}^{t}\right)}\right]}^{\frac{1}{2}}$$

${D}^{t}$ (${x}^{t}$,${y}^{t}$) and ${D}^{t}$(${x}^{t+1}$, ${y}^{t+1}$) denote the distance function of the desired region in period t and period t+1, respectively, with period t serving as the technical reference. ${D}^{t+1}$(${x}^{t}$, ${y}^{t}$) and ${D}^{t+1}$(${x}^{t+1}$,${y}^{t+1}$) have similar meanings. According to the decomposition of the Malmquist index by Fare [42–43], it can be further decomposed into two components: the index of efficiency change (EFFCH), and the index of technological progress change (TECHCH). In addition, with variable returns to scale (VRS) as an assumption, EFFCH can be decomposed into pure efficiency improvement (PECH), and scale efficiency improvement (SECH) index (SECH). If TFPCH>1, it means that the total factor productivity of the test subject has improved during t ~ t + 1. If TFPCH < 1, which indicates a reduction. If TFPCH = 1, this indicates that it remains unchanged. TECHCH > 1 represents a technological advance. PECH represents shift in efficiency due to shift in management level, and PECH > 1 represents an improvement in the management level. SECH > 1 indicates the optimization of the scale efficiency.

2.2.3 Stage 3: Regression analysis of influencing factors

The third stage of the study is to conduct a regression analysis on the panel data to explore the factors influencing efficiency. In this study, the relative efficiency of hospitals in 31 DMUs in mainland China from 2017–2021 was used as the dependent variable (y) and the possible influencing factors were used as independent variables (x).

Determinants of efficiency discussed in the international literature include hospital size, staffing structure, length of stay, location GDP, urbanization rate and government public expenditure, and elderly dependency ratio [42–43]. We selected several influencing factors, but we excluded such indicators because some of the items have a large correlation with others and have a huge impact, such as GDP per capita and location GDP.

Finally, we have collected ten possible influences, including: X1 The proportion of urban population; X2 The proportion of population aged 0–14 and aged 65 and over; X3 The percentage of illiterate population to total aged 15 and over; X4 The level of per capital health expenditure; X5 The level of per capital outpatient and emergency treatment expenditure; X6 The proportion of people covered by health insurance; X7 The percentage of tertiary hospitals to total hospitals; X8 The proportion of hospital personnel; X9 The proportion of primary medical personnel; X10 The proportion of the volume of medical service in primary medical facilities.

X1, X2, and X3 represent the characteristics of each region, reflecting the degree of urbanization, the proportion of disadvantaged groups, and the level of education of the general public; The X4 and X5 reflect the level of health care consumption, which reflects the health care needs of the general public; X6 indirectly indicates the popularity of the basic medical insurance system, which is a very important factor, because in China, local governments believe that it X6 is a very important factor, because in China, local governments believe that it can considerably improve the accessibility and utilization of medical services for all people, which can considerably relieve the medical pressure of residents and is a powerful guarantee for improving people's health; X7 and X8 are representative of the quality of care, and can be viewed as a proxy for the composition of the hospital hierarchy and the distribution of large general hospitals. Secondly, because the most abundant health resources (senior health personnel, advanced medical equipment, etc.) are concentrated in tertiary hospitals, X8 was chosen to represent the human input to the hospital health care delivery system; X9 represents the human input to the establishment of the primary health care system, while X10 represents the output of the primary health care system in terms of health services, which were introduced to assess the impact of the primary health care system on the hospital service system. In addition, we hypothesize that X1, X2, X3, X6, X9 and X10 will have a positive effect on efficiency, while X4, X5, X7 and X8 may have a negative effect on efficiency.

2.3 Statistical and analysis

SPSS 25.0 software was used to collect data related to hospitals in mainland China, and descriptive and comparative analyses were performed on the inputs and outputs of the hospital service systems in each region. The Excel 2019 software was used to process the data, calculate the Gini coefficients for hospitals in different regions for each year and for each dimension, and plot Lorenz curves based on the relevant cumulative percentages in the Gini coefficient calculation process. The efficiency was estimated using MATLEB 2021b and DEAP 2.1 programs, respectively, and the average efficiency was calculated based on the geometric mean of the DEA output of the ultra-efficient. The deterministic analysis of the panel data was performed using STATA 17.0. The significance level was set at 0.05 for all tests in this study and they were all two-sided tests.

3.1 Description of hospital health resource

According to the panel data, from 2017 to 2021, most DMUs show a gradual increase in all inputs, with the exception of some DMUs in specific years. In addition, the average growth rates of the number of hospitals, personnel and beds in mainland China in the last five years were 3.95%, 6.49% and 4.76%, respectively. Shandong has the largest number of hospitals (2654), followed by Sichuan and Henan provinces with 2481 and 2410 respectively, while Ningxia (213) and Tibet (179) have the fewest. There is a wide gap between the provinces with the highest and lowest number of hospitals. While the minimum numbers for Qinghai and Tibet, relatively speaking, as the same provincial unit, are 36359 and 15851 health technicians respectively, a large gap with Shandong; and from the perspective of municipalities directly under the Central Government, Beijing (199262) and Shanghai (153841) have the most hospital health technicians, 2.37 times and 1.83 times more than Tianjin (84169); In terms of the number of hospital-owned beds, Henan and Shandong provinces have the largest number of beds, 536984 and 522326 respectively, far exceeding the average number of beds in mainland Chinese hospital service systems of 182854. The variation of each input for these DMUs is shown in Fig. 1, with significant regional differences between the highest and lowest input indicators (P < 0.05). The ranking of each medical resource is decreasing from East to West, and the gap in the number of hospitals and beds is not large, with slightly more in the East than in the Central and West. While in terms of health technicians, there is a huge gap between the East and the West, with the highest number of 3.1 million in the East and 2.04 million and 1.96 million in the Central and West, respectively, which is related to the peculiar economic and geographical advantages of the East, as shown in Table 1.

Table 1

Investment in medical resources by regional hospital service system (2017–2021)
Variable	Region	2017	2018	2019	2020	2021
Hospital	Total	31056	33009	34354	35394	36570
	East	12219	13036	13445	13816	14252
	Central	8819	9481	10019	10506	11009
	West	10018	10492	10890	11072	11309
Personnel	Total	5784712	6129201	6487497	6774764	7115465
	East	2566583	2710833	2846164	2960380	3107952
	Central	1657976	1758800	1862699	1958763	2046625
	West	1560153	1659568	1778634	1855621	1960888
Beds	Total	6120484	6519749	6866546	7131186	7414228
	East	2495447	2626720	2740018	2845171	2949746
	Central	1879075	2023466	2135332	2246479	2342310
	West	1745962	1869563	1991196	2039536	2122172

3.2 Equity in the allocation of hospital health resources

The Gini coefficient results for 2017–2021 show that the Gini coefficient of the Chinese hospital service system distributed by population decreased from 0.181 to 0.155, with minor fluctuations in the Gini coefficient of the economic perspective and the Gini coefficient of the geographic distribution. By plotting the trend of the Gini coefficients of hospitals in mainland China in the population/economy/geography dimensions from 2017–2021 (Fig. 2), it can be seen that the Gini coefficients of the three dimensions are basically stable at the every level, and the decreasing trend is more obvious in the demographic dimension, which means that the level of resource allocation of hospitals in mainland China has reached the international equity level in terms of population (< 0.2) [13, 15, 29, 44], the Gini coefficient in the geographic dimension hardly fluctuates, but its level in the last five years has remained around 0.615, which is higher than the standard of the alert line (0.4), indicating that the resource allocation level of the hospital service system is more reasonable in the demographic dimension, followed by the economic dimension, which is basically below 0.3.

From the Lorenz curve of hospital medical resource allocation, we can see that the arc of the Lorenz curve of the population dimension is becoming smaller, gradually approaching the absolute equity line, and the residual area between it and the absolute equity line becomes smaller, which indicates a more equitable distribution of resources; the Lorenz curves of the economic dimension and the geographic dimension have mostly no change fluctuations during the five-year period, but there is a significantly different from the Lorenz curves of the population and economic dimensions, which deviated more from the absolute equity line, which again indicates that the allocation of hospital health resources in mainland China is fewer equitable in the geographic dimension [45]. Figure 3 shows the Lorenz plot for each year.

3.3 Hospital Service Efficiency

3.3.1 SE-SBM-DEA

This section is devoted to the study of hospital service efficiency, where the efficiency values of each DMU at different times are calculated based on the SE-SBM-DEA model.

On the whole, the overall efficiency of the hospital service system in mainland China from 2017–2021 is around 0.794, indicating that there is still some room for improving the overall service efficiency of hospitals in China.

According to available literature, the distribution of healthcare resources is uneven between different regions, especially between the East and the West [41, 46]. And most studies point out that the East is generally more efficient than the West, in terms of various types of hospitals and primary care systems, among other things. However, this differs from our results. In terms of zoning, the highest average efficiency was 0.851 in the western region, followed by the eastern region (0.834) and the central region (0.673). In terms of the annual effective unit share, the number of regions with effective operational efficiency is around 13 in 2017–2021, and in 2021, the minimal effective number of hospital service system efficiency is 3 in the eastern region, 4 in the central region, and 8 in the western region (the reference value is minor due to the limited number of hospitals and population in Tibet).

From a macroscopic point of view, the distribution of service efficiency in China's hospital service system shows a relatively stable pattern, mainly showing a gradual increase from north to south, and relatively speaking, the service efficiency in the northern regions (mainly in the northeastern provinces, Shanxi, and Inner Mongolia) has been at a lower level and has fundamentally not fluctuated [40, 47]. From a microscopic point of view, the regions in higher service efficiency still account for a minority of Shanghai, Guangdong, Ningxia, Henan and other regions have been maintaining high service efficiency in the past five years, and in the central and western regions of Guizhou, Hunan, Chongqing and other regions have made significant progress (Table 2).

Table 2

SE-SBM-DEA results by region (2017–2021)
REGION	2017	2018	2019	2020	2021
EAST	0.830	0.865	0.858	0.813	0.807
Beijing	0.676	0.687	1.006	0.599	0.708
Tianjin	0.730	0.747	1.008	0.722	0.767
Hebei	0.490	0.689	0.639	0.681	0.581
Liaoning	0.457	0.544	0.505	0.516	0.465
Shanghai	1.343	1.408	1.388	1.353	1.421
Jiangsu	0.711	0.862	0.840	0.906	0.838
Zhejiang	1.002	1.014	1.003	1.026	1.010
Fujian	0.686	0.711	0.691	0.763	0.662
Shandong	1.025	1.000	0.744	0.811	0.886
Guangdong	1.731	1.189	1.195	1.170	1.188
Hainan	1.005	1.013	0.783	0.748	0.771
CENTRAL	0.590	0.740	0.671	0.679	0.694
Shanxi	0.376	0.396	0.387	0.438	0.400
Jilin	0.504	0.519	0.485	0.455	0.461
Heilongjiang	0.418	0.475	0.465	0.333	0.350
Anhui	0.650	0.839	0.790	0.841	0.773
Jiangxi	0.590	1.002	0.805	1.011	1.015
Henan	1.017	1.060	1.042	1.050	1.032
Hubei	1.025	1.033	1.036	0.751	1.013
Hunan	0.465	1.004	0.684	1.012	1.015
WEST	0.713	0.901	0.888	0.905	0.867
Inner Mongolia	0.454	0.453	0.450	0.474	0.447
Guangxi	1.026	1.018	1.014	1.044	1.025
Chongqing	0.561	0.760	0.781	0.876	1.017
Sichuan	1.026	1.009	1.015	1.019	1.026
Guizhou	0.439	1.008	1.005	1.003	1.004
Yunnan	0.604	1.024	0.784	1.022	1.016
Tibet	2.353	2.179	2.237	2.038	1.984
Shaanxi	0.482	0.592	0.579	0.603	0.573
Gansu	0.588	1.035	0.813	0.781	0.516
Qinghai	0.616	0.741	0.742	1.035	1.003
Ningxia	1.015	1.041	1.057	1.024	1.048
Xinjiang	0.586	0.762	1.018	0.652	0.597

3.3.2 DEA-based Malmquist total factor productivity index

After evaluating the efficiency values for each year, we use the Malmquist index to decompose the total factor productivity of the hospital service system in mainland China to further explore the dynamic evolution of efficiency and the deep-seated causes that influence the change in efficiency. Tables 3 and 4 show the TFPCH and its decomposition in the temporal and geographic dimensions, respectively.

As can be seen from Table 3, the five-year average TFPCH declined in all 31 regions during the study period, with all provinces showing varying degrees of decline, with the smallest declines in Zhejiang, Shanghai, Chongqing, 0.8%, 2.1% and 2.3%, respectively; with the largest decline of 15.1% per year in Guangdong (0.849), whose TFPCH is mainly limited by the TECHCH. The TECHCH index declined in 16 regions, including Zhejiang, Shanghai, Chongqing and Beijing, while the EFFCH index was all > 1.

As can be seen through Table 4, the average TFPCH of the hospital service system is less than the standard level 1 (0.955), and decreases at a rate of 4.5% per year. It indicates that the overall efficiency of China's hospital service system is not satisfactory, and the decline trend and characteristics have not improved significantly. The highest TFPCH was observed in 2020–2021 (1.054), and the most significant decline was observed in 2019–2020 (0.820). Improvements in the efficiency of health service systems can be attributed to technological advances, improvements in human capital, and other environmental factors. From the decomposition metrics, excluding 2020–2021 when both efficiency metrics exceed 1, both EFFCH and TECHCH have inefficiency problems in the remaining years, and both EFFCH and TECHCH have efficiency metrics below 1 in 2019–2020, which jointly constrain the improvement of TFPCH and the effect of TECHCH is larger. The decomposition of EFFCH shows that both effects are tiny and the difference is not significant.

Table 3

MTFP and its decomposition (region)
DMU	PECH	SECH	EFFCH	TECHCH	TFPCH
Beijing	1.004	0.998	1.002	0.973	0.975
Tianjin	1.003	1.001	1.005	0.965	0.969
Hebei	0.965	1.004	0.968	0.973	0.942
Liaoning	0.993	1.000	0.993	0.951	0.944
Shanghai	1.000	1.000	1.000	0.979	0.979
Jiangsu	1.003	0.994	0.997	0.966	0.963
Zhejiang	1.000	1.012	1.012	0.980	0.992
Fujian	0.989	0.999	0.988	0.973	0.962
Shandong	0.998	1.001	0.999	0.973	0.972
Guangdong	1.000	0.988	0.988	0.859	0.849
Hainan	0.983	0.995	0.979	0.950	0.930
Shanxi	1.000	0.999	0.999	0.971	0.970
Jilin	0.981	1.000	0.981	0.950	0.932
Heilongjiang	0.989	0.997	0.985	0.939	0.925
Anhui	0.984	1.002	0.987	0.964	0.952
Jiangxi	1.015	1.000	1.015	0.961	0.975
Henan	1.000	1.011	1.011	0.962	0.972
Hubei	1.000	1.010	1.010	0.959	0.968
Hunan	1.021	1.007	1.029	0.941	0.968
Inner Mongolia	0.982	0.998	0.980	0.970	0.951
Guangxi	1.000	1.000	1.000	0.972	0.972
Chongqing	1.026	1.001	1.027	0.951	0.977
Sichuan	1.000	1.027	1.027	0.945	0.970
Guizhou	1.018	1.000	1.019	0.952	0.970
Yunnan	1.008	1.002	1.010	0.966	0.975
Tibet	1.000	1.000	1.000	0.895	0.895
Shaanxi	0.987	1.002	0.990	0.972	0.962
Gansu	0.989	1.001	0.990	0.926	0.917
Qinghai	1.030	1.003	1.034	0.938	0.970
Ningxia	1.000	1.000	1.000	0.958	0.958
Xinjiang	0.996	1.001	0.997	0.957	0.955

Table 4

MTFP and its decomposition (year)
Year	PECH	SECH	EFFCH	TECHCH	TFPCH
2017–2018	1.024	1.003	1.027	0.942	0.967
2018–2019	0.988	1.006	0.995	1.000	0.995
2019–2020	0.965	0.999	0.964	0.850	0.820
2020–2021	1.019	0.999	1.018	1.035	1.054
Mean	0.999	1.002	1.001	0.954	0.955

3.4 Regression analysis

In this study, based on literature studies and speculation, factors that may affect the efficiency of the hospital service system in each province are introduced into the regression analysis model of the panel data to explore their role. We provide detailed data for these 10 independent variables in Supplementary File 1.

We performed regression analysis of panel data using fixed effects regression (fe) and random effects regression (re) with BP test and the results were significant. Finally, using the Hausman test, the results showed that P < 0.05, the original hypothesis was rejected and the use of fixed effects regression analysis was more appropriate. After the results of the panel data regression analysis showed that the average efficiency of the 31 DMUs in 2017–2021 was 0.857 ± 0.352. The detailed results of the regression analysis are shown in Table 5, in which five of the ten variables had no significant effect on the service efficiency of the hospital service system, including X2 (The proportion of population aged 0–14 and aged 65 and over, P = 0.645), X4 (The level of per capital health expenditure, P = 0.057), X6 (The proportion of people covered by health insurance, P = 0.196), X7 (The percentage of tertiary hospitals to total hospitals, P = 0.070), and X9 (The proportion of primary medical personnel, P = 0.257). Five variables have a significant effect on the service efficiency of the hospital service system, including: X1 (The proportion of urban population, P = 0.000), X3 (The percentage of illiterate population to total aged 15 and over, P = 0.021), and X10 (The proportion of the volume of medical service in primary medical facilities, P = 0.026), indicate that an increase in the proportion of the urban population, the percentage of illiterate population to total aged 15 and over, and the proportion of the volume of medical service in primary medical facilities will improve the efficiency score of the hospital service system. X5 (The level of per capital outpatient and emergency treatment expenditure, P = 0.001), X8 (The proportion of hospital personnel, P = 0.000) were negatively correlated with hospital service system, indicating that the level of per capital outpatient and emergency treatment expenditure and the proportion of hospital personnel will limit the efficiency score of provincial hospital service system. The panel data regressions in this study do not take the degree of contribution into account, so it is not possible to analyze the degree of impact of each significant influence.

Table 5

Results of regression analysis of influencing factors
Variable	Non-standardized coefficients		t	P- values	95% CI	Interval
Variable	B	SD	t	P- values	95% CI	Interval
X1	0.056	0.014	3.930	0.000	0.028	0.085
X2	0.005	0.010	0.460	0.645	-0.016	0.025
X3	0.029	0.012	2.330	0.021	0.004	0.054
X4	0.039	0.020	1.930	0.057	-0.001	0.080
X5	-0.002	0.001	-3.310	0.001	-0.003	-0.001
X6	0.030	0.023	1.300	0.196	-0.016	0.075
X7	-0.026	0.014	-1.830	0.070	-0.055	0.002
X8	-0.013	0.003	-4.560	0.000	-0.018	-0.007
X9	0.012	0.011	1.140	0.257	-0.009	0.034
X10	0.011	0.005	2.250	0.026	0.001	0.021

The value of α (the constant item) was not presented in this table because each DMU had a unique constant value in the regression model.

Since the new medical reform, China's investment in healthcare has steadily increased, with national health expenditure rising from 175.5 billion RMB in 2009 to 525.9 billion RMB in 2017, and to 768.4 billion RMB in 2021, with an average annual growth rate of 13.10%, from 6.32–6.72% of China's GDP. The scale of investment in health care resources in China's hospital service system has increased significantly under the high attention of the state [40–41], and the number of hospitals owned by China is 36,570 in 2021, an increase of 5,514 in five years.

In 2009, the Chinese government proposed to optimize the allocation of health resources and improve the efficiency of health resource utilization [48]. As China's medical reform progresses, governments at all levels have attempted to address these issues through a series of actions such as rationalizing the allocation of health resources and improving the efficiency of utilization. However, most of the series of documents issued by the Chinese government to optimize the allocation of health resources have been based on population allocation. Accordingly, the equity of allocating resources based on population size is much better than the equity of allocating resources based on geographic and economy, as verified by the results of this study [49]. But because of China's vast geographical area, geographical inequality remains an unavoidable problem in resource allocation. Other developing countries, such as Mexico and Vietnam, also have significant inequalities in the geographic distribution of health care resources. In China, capital resources are concentrated in the developed eastern provinces, which are significantly richer in the East, while the Central and West of the country lag relatively behind in development due to a lack of priority and motivation [13, 50].

The results for the Gini coefficient and Lorentz curve show that the Lorentz curve of the hospital service system is closest to the absolute equity line in the demographic dimension, followed by the economic perspective and the geographic perspective. The results for the Gini coefficients show that the Gini coefficients of the geographic and economic dimensions are less volatile, and the Gini coefficient of the demographic dimension is continuously decreasing. Specifically, the equity ranking of hospital health resources in each dimension is population/economy/geography. This finding is consistent with other studies on the equity of hospital health resource allocation in China [44, 51–52]. There are two main reasons for this result: on the one hand, the government issues policies primarily to meet the medical needs of the population, rather than the regional layout; on the other hand, there are disparities in the level of economic development in different provinces, which has an impact on the supply of medical resources.

The Malmquist index model measures the change in dynamical efficiency of DMUs over multiple periods. It shows that the overall efficiency of the mainland's hospital service system is relatively flat, remaining at around 0.955. The findings showed that all 31 DMUs have total factor productivity levels below 1, with all provinces showing varying degrees of decline, possibly related to the input and output indicators chosen for this study, as well as the fact that mainland China was hit by the COVID-19 in 2020, which had a greater impact on the efficiency of hospital services [38, 53]. The regional average productivity index pattern shows minor differences between the eastern, central and western regions, and the overall efficiency of China's hospital service system is unsatisfactory, with no significant improvement in the decline trend and characteristics. The results of the SE-SBM-DEA calculations indicate that the number of DMUs with effective efficiency does not exceed half per year. The efficiency of hospital services varies significantly from region to region. This finding is similar to the findings of Ben [54]. The efficiency scores varied across regions, with some regions performing very well (e.g., Tibet, Shanghai, and Guangdong), while several others barely reached 0.5 (e.g., Heilongjiang, Jilin, and Liaoning). Tibet, located in western China, has a low population density, low per capita GDP and an underdeveloped economy. Its regional economic development level limits the influx of government investment, medical institution construction, and medical personnel to the western region [55]. Shanghai and Guangdong, located in the developed eastern region, have inherent advantages in terms of medical institution construction, number of health technicians and health management. Greater government investment in health care, reasonable growth in inputs and outputs, focus on technological shifts and increased management efficiency have maintained elevated levels of efficiency in hospital services in the two provinces. However, the northeast lags behind in terms of human resources and financial investment, which, along with relatively inefficient management models and high-tech applications, have contributed to the low levels of efficiency in the three northeastern provinces. In terms of sub-region, hospital efficiency is ranked as West > East > Central. In fact, this result is not entirely unexpected, as the Chinese government has been committed to improving health care development and health outcomes in western China through the implementation of various priority policies in the western region, in contrast to the substantial redundancy of inputs that have been underutilized in the eastern region. Compared with the West, the Central has not benefited from preferential regional policies, making it the most vulnerable region. Meanwhile, the western region has a limited level of technology, but relatively appropriate and fully utilized inputs [56–57].

The determinants of efficiency are complex and the results of the regression analysis of the panel data are not entirely consistent with our initial assumptions. In this study, regression results showed that the effect of disadvantaged populations on the efficiency of hospital services was not significant, possibly because more disadvantaged people were attending primary care facilities due to the implementation of triage systems; X3(The percentage of illiterate population to total aged 15 and over) has a positive effect on the operational efficiency of the provincial hospital service system, which means that the increase of X3(The percentage of illiterate population to total aged 15 and over) increases the demand for hospital medical services; X10(The proportion of the volume of medical service in primary medical facilities) has a positive effect relationship with service efficiency, which means that an increase in the proportion of primary medical institution services will increase the efficiency score of the provincial hospital service system; X1(The proportion of urban population) reflects the potential demand for medical services in the area, which will affect the efficiency of the hospital. While X5 and X8 have a negative relationship with the efficiency of hospital services, this means that an increase in per capita healthcare costs and the ratio of hospital medical staff to all medical staff will limit the efficiency score of the provincial hospital service system.

Efficiency reflects the relationship between the inputs and outputs of the hospital service system. Based on the allocation of health resources and the relative efficiency scores of provincial hospital service systems, we recommend that: 1 More attention be paid to "improving outputs" rather than "increasing inputs" in the operation of the hospital service system. The lack of technical and human resources for health has led to modest improvements in overall efficiency as hospitals have grown in size. 2 For the central and western regions, provinces with low TE should focus on improving their management capacity and technical level. For provinces with low SE, they should increase the capital investment according to the actual situation of hospitals, but should avoid "rough" investment and expand the scale of hospitals in this region reasonably and effectively. 3 In different provinces, local governments should implement different programs based on their own medical and health resource allocation levels, preference of residents for medical treatment and health needs. 4 Focus on innovative development, promote the output of modern technologies, new products and improved strategies, introduce new technologies and management concepts, such as health technology assessment, and explore more scientific and rational paths for hospital development.

To the best of our knowledge, this is one of the rare large-scale longitudinal analyses using representative provincial data to investigate equity in hospital health resource allocation and health service efficiency, and to explore the factors that influence service efficiency. There are several limitations to this study. The study did not distinguish between hospitals for differences in service quality. Moreover, due to the limitations in the selection of DEA indicators, the selection of input and output indicators is limited to data that have been counted and cannot be calculated from a more detailed perspective. Furthermore, this study assesses the equity of hospital health resource allocation from a provincial perspective in terms of demographic, economic and geographic aspects, and does not discuss inequalities within provinces, across populations, etc. How to optimize the allocation of health resources requires further, more comprehensive research. This may be the next topic of our study.

a. Ethics approval and consent to participate

Not applicable. The Data used in this study is publicly available.

b. Consent for publication

Not applicable.

c. Availability of data and materials

The data used in this article are from the China Statistical Yearbook and the China Health Care Statistical Yearbook. If anyone wants to request the data from this study, please contact Ruxin Kou (E-mail: [email protected]).

d. Competing interests

The authors declare that they have no competing interests.

e. Funding

National Natural Science Foundation of China (71774119); Soft Science Research Program of Shandong (2022RKY07011)

f. Authors’ contributions

Ruxin Kou wrote the first draft of the paper; Kangni Mei collected and analyzed relevant literature. Yuqing Bi and Jingwen Huang were responsible for the construction of the logic and framework of the article; Shilan Yang and Kexuan Chen led the collection and analysis of data metrics; Wei Li was responsible for the theoretical study and the final data results. All authors have read and approved the final draft. The corresponding author is Wei Li.

g. Acknowledgments

We thank all the authors for their work and support in this study and for their valuable comments and suggestions.

h. Authors' information

1 School of Public Health, Weifang Medical University, Weifang 261021, Shandong, China

Ruxin Kou, E-mail: [email protected]

Kangni Mei, E-mail: icyconnie520@163

Yuqing Bi, E-mail: [email protected]

Jingwen Huang, E-mail: [email protected]

Shilan Yang, E-mail: [email protected]

Kexuan Chen, E-mail: [email protected]

Wei Li, E-mail: [email protected]

Dohmen P, van Ineveld M, Markus A et al. Does competition improve hospital performance: a DEA based evaluation from the Netherlands[J].The European Journal of Health Economics, 2022:1–19.
Piubello Orsini L, Leardini C, Vernizzi S, et al. Inefficiency of public hospitals: a multistage data envelopment analysis in an Italian region[J]. BMC Health Serv Res. 2021;21(1):1–15.
İlgün G, Sönmez S, Konca M, et al. Measuring the efficiency of Turkish maternal and child health hospitals: A two-stage data envelopment analysis[J]. Eval Program Plan. 2022;91:102023.
Mohamadi E, Kiani MM, Olyaeemanesh A et al. Two-step estimation of the impact of contextual variables on technical efficiency of hospitals: the case study of public hospitals in Iran[J].Frontiers in Public Health, 2022:1839.
Fumbwe F, Lihawa R, Andrew F, et al. Examination on level of scale efficiency in public hospitals in Tanzania[J]. Cost Eff Resource Allocation. 2021;19(1):1–10.
Zakowska I, Godycki-Cwirko M. Data envelopment analysis applications in primary health care: a systematic review[J]. Fam Pract. 2020;37(2):147–53.
Mirmozaffari M, Yazdani R, Shadkam E, et al. A novel hybrid parametric and non-parametric optimisation model for average technical efficiency assessment in public hospitals during and post-COVID-19 pandemic[J]. Bioengineering. 2021;9(1):7.
Huang H, Zhong L, Shen T, et al. Performance prediction and optimization for healthcare enterprises in the context of the COVID-19 pandemic: an intelligent DEA-SVM model[J]. J Comb Optim. 2022;44(5):3778–91.
Cheng Z, Cai M, Tao H, et al. Efficiency and productivity measurement of rural township hospitals in China: a bootstrapping data envelopment analysis[J]. BMJ open. 2016;6(11):e011911.
Guo X, Zhang J, Xu Z, et al. The efficiency of provincial government health care expenditure after China’s new health care reform[J]. PLoS ONE. 2021;16(10):e0258274.
Zhongming Z, Linong L, Xiaona Y et al. The 14th Five-Year Plan of the People’s Republic of China—Fostering High-Quality Development[J]. 2021.
Tao Y, Henry K, Zou Q, et al. Methods for measuring horizontal equity in health resource allocation: a comparative study[J]. Health Econ Rev. 2014;4:1–10.
Zhang Y, Wang Q, Jiang T, et al. Equity and efficiency of primary health care resource allocation in mainland China[J]. Int J Equity Health. 2018;17:1–12.
Toyabe S. Trend in geographic distribution of physicians in Japan[J]. Int J Equity Health. 2009;8(1):1–8.
Ding J, Hu X, Zhang X, et al. Equity and efficiency of medical service systems at the provincial level of China’s mainland: a comparative study from 2009 to 2014[J]. BMC Public Health. 2018;18:1–14.
Ahmed S, Hasan MZ, MacLennan M, et al. Measuring the efficiency of health systems in Asia: a data envelopment analysis[J]. BMJ open. 2019;9(3):e022155.
Andrews A. Investigating technical efficiency and its determinants: The Case of New Zealand District Health Boards[J]. Health Policy and Technology. 2020;9(3):323–34.
Gattoufi S, Oral M, Reisman A. A taxonomy for data envelopment analysis[J]. Socio-Economic Plann Sci. 2004;38(2–3):141–58.
Sherman HD. Hospital efficiency measurement and evaluation: empirical test of a new technique[J]. Med Care. 1984;22(10):922–38.
Charnes A, Cooper WW, Rhodes E. Evaluating program and managerial efficiency: an application of data envelopment analysis to program follow through[J]. Manage Sci. 1981;27(6):668–97.
Banker RD, Charnes A, Cooper WW et al. A comparison of DEA and translog estimates of production frontiers using simulated observations from a known technology[J].Applications of modern production theory: Efficiency and productivity, 1988:33–55.
Zhao T. Efficiency of Institutional Health Care Service in Rural China Taking Account of Demand-Side Factors: Empirical Study Based on the Survey of Three Provinces. Nankai Econ[J] Stud. 2017;3:34–49.
Zheng D, Gong J, Zhang C. Efficiency of medical service systems in the rural areas of Mainland China: A comparative study from 2013 to 2017[J]. Public Health. 2019;171:139–47.
Du T, Ran L, Li JL. Dynamic evaluation of medical and health service efficiency in China based on DtSBM model[J]. J Beijing Inst Technol (Soc Sci Ed). 2019;21:82–94.
Yu JL, Qian ZW. Investment gap in health care industry and the temporal-spatial evolution of health service efficiency at provincial administrative institutions of China[J]. J Arid Land Resour Environ. 2018;32:59–66.
Hou Y, Tao W, Hou S et al. Levels, trends, and determinants of effectiveness on the hierarchical medical system in China: Data envelopment analysis and bootstrapping truncated regression analysis[J].Frontiers in Public Health, 2022,10.
Mohamadi E, Manesh AO, Takian A, et al. Technical efficiency in health production: A comparison between Iran and other upper middle-income countries[J]. Health Policy and Technology. 2020;9(3):335–47.
Giménez V, Prieto W, Prior D, et al. Evaluation of efficiency in Colombian hospitals: An analysis for the post-reform period[J]. Socio-Economic Plann Sci. 2019;65:20–35.
Yu H, Yu S, He D, et al. Equity analysis of Chinese physician allocation based on Gini coefficient and Theil index[J]. BMC Health Serv Res. 2021;21(1):1–8.
Yang L, Cheng J. Analysis of inequality in the distribution of general practitioners in China: evidence from 2012 to 2018[J]. Volume 23. Primary Health Care Research & Development; 2022. p. e59.
Li Q, Wei J, Jiang F, et al. Equity and efficiency of health care resource allocation in Jiangsu Province, China[J]. Int J Equity Health. 2020;19:1–13.
Bahrami MA, Rafiei S, Abedi M, et al. Data envelopment analysis for estimating efficiency of intensive care units: a case study in Iran[J]. International journal of health care quality assurance; 2018.
Huang M, Luo D, Wang Z, et al. Equity and efficiency of maternal and child health resources allocation in Hunan Province, China[J]. BMC Health Serv Res. 2020;20(1):1–10.
Jing R, Xu T, Lai X, et al. Technical efficiency of public and private hospitals in Beijing, China: a comparative study[J]. Int J Environ Res Public Health. 2020;17(1):82.
Chen Z, Chen X, Gan X, et al. Technical efficiency of regional public hospitals in China based on the three-stage DEA[J]. Int J Environ Res Public Health. 2020;17(24):9383.
Li Z, Yang L, Tang S, et al. Equity and efficiency of health resource allocation of Chinese Medicine in mainland China: 2013–2017[J]. Front Public Health. 2020;8:579269.
Sultan WIM, Crispim J. Are public hospitals reforming efficiently in West Bank?[J]. Confl Health. 2018;12(1):1–14.
Medarević A, Vuković D. Efficiency and Productivity of Public Hospitals in Serbia Using DEA-Malmquist Model and Tobit Regression Model, 2015–2019[J]. Int J Environ Res Public Health. 2021;18(23):12475.
Alatawi AD, Niessen LW, Khan JAM. Efficiency evaluation of public hospitals in Saudi Arabia: an application of data envelopment analysis[J]. BMJ open. 2020;10(1):e031924.
Tone K. A slacks-based measure of efficiency in data envelopment analysis[J]. Eur J Oper Res. 2001;130(3):498–509.
Du J, Liang L, Zhu J. A slacks-based measure of super-efficiency in data envelopment analysis: A comment[J]. Eur J Oper Res. 2010;204(3):694–7.
Ali M, Debela M, Bamud T. Technical efficiency of selected hospitals in Eastern Ethiopia[J]. Health Econ Rev. 2017;7:1–13.
Campanella P, Azzolini E, Izzi A, et al. Hospital efficiency: how to spend less maintaining quality?[J]. Annali dell'Istituto superiore di sanita. 2017;53(1):46–53.
Jian JIN, Jianxiang W, Xiaoyi MA, et al. Equality of medical health resource allocation in China based on the Gini coefficient method[J]. Iran J public health. 2015;44(4):445.
Wu J, Yang Y. Inequality trends in the demographic and geographic distribution of health care. Int J Health Plann Manag. 2019;34(1):e487–e508. professionals in China: Data from 2002 to 2016[J].
Kiadaliri AA, Jafari M, Gerdtham UG. Frontier-based techniques in measuring hospital efficiency in Iran: a systematic review and meta-regression analysis[J]. BMC Health Serv Res. 2013;13:1–11.
Tran TH, Mao Y, Nathanail P, et al. Integrating slacks-based measure of efficiency and super-efficiency in data envelopment analysis[J]. Omega. 2019;85:156–65.
Wei JM, Gao QL, Huang MZ, et al. Study on the operation efficiency of county medical community based on DEA Model[J]. Chin J Health Policy. 2021;2:23–7.
Ran Y, Gao H, Han D et al. Medical Alliances in a County:a Longitudinal Study on a healthcare reform in Rural China[J].
Liu W, Liu Y, Twum P, et al. National equity of health resource allocation in China: data from 2009 to 2013[J]. Int J Equity Health. 2016;15(1):1–8.
Wang Y, Li Y, Qin S, et al. The disequilibrium in the distribution of the primary health workforce among eight economic regions and between rural and urban areas in China[J]. Int J Equity Health. 2020;19(1):1–10.
Wang S, Xu J, Jiang X, et al. Trends in health resource disparities in primary health care institutions in Liaoning Province in Northeast China[J]. Int J Equity Health. 2018;17(1):1–8.
Androutsou L, Kokkinos M, Latsou D, et al. Assessing the Efficiency and Productivity of the Hospital Clinics on the Island of Rhodes during the COVID-19 Pandemic[J]. Int J Environ Res Public Health. 2022;19(23):15640.
Ben H, Zhang Y. Research on the Influence of Social Networks on the Allocation Efficiency of Public Medical Resources—Taking Jiangsu as an Example[J]. Jiangxi Soc Sci. 2018;38:227–34.
Peykani P, Seyed Esmaeili FS, Rostamy-Malkhalifeh M, et al. Measuring productivity changes of hospitals in Tehran: the fuzzy Malmquist productivity index[J]. Int J Hosp Res. 2018;7(3):1–16.
Lin L, Wu F, Chen W, et al. Research on urban medical and health services efficiency and its spatial correlation in China: based on panel data of 13 cities in Jiangsu Province[C]//Healthcare. MDPI. 2021;9(9):1167.
Su W, Du L, Fan Y, et al. Equity and efficiency of public hospitals’ health resource allocation in Guangdong Province, China[J]. Int J Equity Health. 2022;21(1):1–18.

No competing interests reported.

SupplementaryFile1.xlsx

Download PDF

Version 1

posted

You are reading this latest preprint version

Efficiency and equity of hospital services in mainland China: based on data from 2017 to 2021

Status:

Version 1

Abstract

Figures

1. Background

2. Methods

2.1 Data source

2.2 Research Design

2.2.1 Stage 1: Equity of hospital health resource allocation

2.2.2 Stage 2: hospital service efficiency evaluation

2.2.3 Stage 3: Regression analysis of influencing factors

2.3 Statistical and analysis

3. Results

3.1 Description of hospital health resource

3.2 Equity in the allocation of hospital health resources

3.3 Hospital Service Efficiency

3.3.1 SE-SBM-DEA

3.3.2 DEA-based Malmquist total factor productivity index

3.4 Regression analysis

4. Discussion

5. Conclusion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1