Abstract
Background
Against the backdrop of the coordinated advancement of the “Healthy China” and “Digital China” initiatives, enhancing the utilization efficiency of healthcare resources has become a core issue in achieving universal health coverage. Currently, provinces across China face significant challenges in allocating medical resources. Conducting in-depth research into the current state of resource utilization efficiency and its underlying mechanisms holds critical practical significance for optimizing resource allocation and driving high-quality development of the healthcare system.
Methods
This study employs a data envelopment analysis model to measure the efficiency of healthcare resource utilization at the provincial level in China. Utilizing a fuzzy set qualitative comparative analysis method, it systematically examines the synergistic effects of different antecedent conditions to reveal the diverse pathways driving high resource utilization efficiency.
Results
Data analysis reveals a complex picture of healthcare resource utilization efficiency in China: (1) In 2021, China’s average comprehensive healthcare resource utilization efficiency stood at 0.918. However, only 35.5% of provinces achieved DEA efficiency, indicating that nearly two-thirds of provinces still face resource misallocation issues. (2) Among these, 14 provinces exhibited increasing returns to scale (under-investment in resources), while 6 provinces showed decreasing returns to scale (over-investment in resources), reflecting structural imbalances between resource allocation and actual demand. (3) fsQCA configuration analysis identified five efficient driving pathways: H1(Digital-Economic-Spatial Constraint Type) reflects cumulative disadvantages across multiple dimensions; H2 (Economy-Digital Synergy) and H4 (Urbanization-Driven) demonstrate how different factor combinations achieve functional equivalence through divergent pathways; H3 (Comprehensive Factor Balance) and H5 (Full Factor Empowerment) collectively outline the evolutionary path from “factor-driven” to “innovation-driven” development.
Conclusion
Achieving high-level efficiency in healthcare resource utilization does not rely on a single optimization approach, but rather results from the coordinated allocation of economic capital, digital technology, and spatial structures. Research reveals the existence of a “multiple concurrent” equivalent driving model, offering diverse pathways for regions with varying developmental conditions. Policy formulation should abandon a one-size-fits-all approach and instead adopt differentiated strategies aligned with local resource endowments and developmental stages. This systematic thinking will drive comprehensive improvements in healthcare resource utilization efficiency.
Clinical trial number
Not applicable.
Keywords: Economic capital, Digital technology, Spatial structure, Health performance, Health resource utilization , TOE framework
Background
Improving the efficiency of health resource utilization to achieve optimal health outcomes with limited inputs has emerged as a central challenge in global health system governance. This issue pertains not only to whether individuals can equitably access essential medical [1] and health services and exercise their fundamental right to health, but also plays a pivotal role in advancing universal health coverage. Currently, factors such as global population aging, the epidemiological transition toward chronic diseases, and increasingly diverse health needs are collectively shifting the drivers of health system performance—from reliance on singular resource inputs toward a complex, synergistic paradigm that emphasizes the integration of multiple factors.
Against the backdrop of the integrated “Healthy China 2030” strategy and the principle of high-quality development, the core mission of China’s public health system is undergoing a profound transformation—from prioritizing scale expansion to emphasizing efficiency enhancement. In this context, the efficiency of health resource utilization—defined as “how to achieve optimal health outcomes through efficient resource allocation”—has emerged as a central focus in both academic research and policy implementation. As the world’s most populous nation with substantial regional disparities in development, China exhibits considerable heterogeneity in the efficiency of health resource utilization across provinces. Generally, eastern provinces—supported by strong economic foundations and advanced digital technologies—consistently demonstrate superior health performance. In contrast, central and western regions, constrained by weaker economies, underdeveloped digital infrastructure, and suboptimal spatial distribution of resources, encounter significant challenges in efficiently converting health resource inputs into tangible health outputs [2].
Though existing research has extensively examined the determinants of health performance, significant limitations persist. First, the majority of studies rely on conventional quantitative approaches—such as linear regression—emphasizing the “net effect” of individual factors, including economic capital, digital technology, or spatial structure, while neglecting the underlying “causal complexity” in shaping health outcomes. In reality, multiple factors may jointly influence health performance through distinct configurational pathways, reflecting both “multiple conjunctural causation” and “equifinality.” Second, current literature typically treats these three key factors as independent variables, with insufficient attention paid to their interactive mechanisms. For instance, how digital technology might mitigate health disparities arising from inadequate economic capital or spatial imbalances, or how economic capital facilitates the deployment of digital technology and the optimization of spatial configurations, remains underexplored. Finally, there is a notable paucity of configurational studies focusing on the efficiency of provincial health resource utilization in China. This gap hinders the identification of differentiated strategies for enhancing health performance across regions with diverse resource endowments, thereby limiting the precision and policy relevance of existing recommendations.
Literature review
Health resource utilization efficiency based on toe framework
The TOE framework, serving as a systematic theoretical lens for analyzing organizational technology adoption and performance enhancement, integrates internal and external factors as well as technical characteristics within application contexts by categorizing them into three dimensions: technology (T), organization (O), and environment (E) [3]. This framework provides an analytical perspective for understanding the complex mechanisms underlying the efficiency of provincial health resource utilization. The improvement of the utilization efficiency of provincial health resources is essentially a complex process in which the system effectively absorbs, transforms and governs digital technology under the structural constraints of the specific spatial environment and relying on its own economic capital conditions, so as to realize the optimization of resource allocation. Based on this, economic capital, digital technology and spatial structure are respectively corresponding to the dimensions of organization, technology and environment in the toe framework, with interdisciplinary theoretical support. The economic capital of organization dimension is the material basis of system operation and change. The health production function theory shows that the “resource inventory” composed of public finance and residents’ ability to pay directly determines the potential of the system for technology investment and service optimization. Digital technology in the technical dimension is the core enabling tool to drive process reengineering, and the key to its effectiveness lies in the adaptability of technical characteristics and organizational processes. The environmental dimension focuses on spatial structure. Medical geography and institutional theory reveal that population distribution, urban-rural pattern and geographical conditions set the rigid external situation of resource allocation by shaping service accessibility, demand density and supply cost. The toe framework reveals the dynamic synergy among technology, organization and environment: the introduction of technology depends on the support of organizational resources, and the release of efficiency requires the adaptation of environmental scenarios. This integrated perspective goes beyond the common “single factor determinism” in existing studies, and lays a theoretical foundation for further exploring the “configuration effect” of multiple antecedents coupling and analyzing the differentiated path of China’s provincial health resource utilization efficiency.
Economic capital and the efficiency of health resource utilization
The organizational dimension within the OE framework emphasizes the internal resource endowments and capabilities of an organization, encompassing its scale structure, human resources, and available assets. These elements collectively determine the organization’s potential to identify, assimilate, and leverage technology to enhance performance. According to the resource-based view, economic capital constitutes the most fundamental input for health services. The health demand model lays down the classic rationale for how income influences the efficiency of accessing and utilizing medical resources through budget constraints [4]. Mainstream research has validated its “scale effect,” as evidenced by OECD country data indicating a positive correlation between economic level and resource utilization efficiency [5]. However, empirical studies within the Chinese context have unveiled the “conditionality” and “structural nature” of economic capital’s role, underscoring the necessity for an in-depth analysis considering it as an organizational condition. Firstly, notable regional heterogeneity exists. The marginal contribution of economic capital to health resource utilization efficiency is significantly higher in eastern provinces than in western ones [6]. The study highlights a “structural mismatch” in economic investment in the west, where capital is overly concentrated on hardware rather than human resources and service processes, resulting in diminishing or even negative marginal returns on capital investment [7]. Secondly, its impact is modulated by the internal institutional arrangements of organizations. The “double-edged sword effect” of residents’ payment capacity suggests that, in the absence of a comprehensive medical insurance system serving as a collaborative payment mechanism, medical expense inflation could erode the efficiency gains derived from private income growth [8].
Digital technology and health resource utilization efficiency
Within the TOE framework, the technology dimension centers on the inherent features of technology and its compatibility with organizational structures and processes. As an empowering tool, digital technology can lower transaction costs by enhancing information symmetry, improving service accessibility, and optimizing processes, thereby boosting the marginal output of resources. The World Health Organization emphasizes in its Global Strategy on Digital Health (2020–2025) that digital tools can effectively reduce transaction costs within healthcare systems by promoting information symmetry, enhancing service accessibility, and optimizing operational processes, thereby increasing the marginal output of resources [9, 10]. Research findings have indicated that an increase of one unit in a region’s digital health index is associated with a substantial enhancement in the overall productivity of its healthcare system. This enhancement is more pronounced in areas that are characterised by a paucity of medical resources. This preliminary finding lends support to the hypothesis that the digital dividend may possess a “universal” character [11, 12]. However, the essence of the TOE framework lies in highlighting that the realization of technology’s potential value is profoundly constrained by organizational and environmental conditions, thereby giving rise to complex “boundary conditions. Research grounded in the Chinese context offers abundant evidence to support this. Firstly, governance capacity serves as a prerequisite for technological empowerment. The research reveals that the positive impact of internet healthcare platform usage on the efficiency of health resource utilization is magnified by 1.7 times in provinces with high fiscal transparency, whereas it is nearly negligible in regions plagued by severe data silos, indicating that “governance absorption” serves as a crucial prerequisite for technological empowerment [13, 14]. Secondly, environmental disparities result in an uneven distribution of technological benefits. Regionally, a notable spatial correlation exists between the provincial variations in China’s digital infrastructure and the disparities in health resource utilization efficiency [15, 16]. From a demographic perspective, the “ability gap” in digital health literacy among the elderly, low-income individuals, and those with lower education levels, coupled with “institutional barriers” related to data privacy and interoperability, collectively pose significant obstacles to the comprehensive application of technology [17–19]. These findings collectively demonstrate that the influence of digital technology is neither independent nor uniform; rather, it is intricately linked to regional economic capital and spatial structures, potentially exacerbating existing resource disparities [20–22].
Spatial structure and health resource utilization efficiency
In the TOE theoretical framework, the environmental dimension refers to the external context in which an organization operates, encompassing the market, external support, and community pressures [23]. For health systems, spatial structures encompass population distribution, geographic distances, and urbanization patterns, constituting their most fundamental environmental context. The impact of spatial structure on the efficiency of health resource utilization originated from research on “geographical accessibility“ [24]. Goddard and Smith, [25] pointed out that for each additional kilometer patients must travel for medical services, the likelihood of seeking care drops by 0.6%–1.2%, indicating that spatial layout directly constrains the effective conversion of resources into healthy outputs. Relevant research has substantiated that spatial agglomeration enhances efficiency through economies of scale and knowledge spillover effects. From the theoretical perspective of agglomeration effects, high population density and urbanization rates improve efficiency through three main pathways: “economies of scale,” “technological spillover,” and “deepening division of labor“ [26]. The spatial concentration of healthcare resources reduces equipment idleness; multidisciplinary collaboration improves the quality of diagnosis and treatment; and specialized division of labor optimizes the structure of resource allocation [27].Empirical studies in China’s Yangtze River Delta show that for every 10% increase in urbanization rate, health resource utilization efficiency (measured by DEA comprehensive efficiency) increases by 8.2%. The marginal contribution of population density follows an “inverted U-shape,” with optimal efficiency in the range of 600–800 people per square kilometer [28, 29]. However, spatial agglomeration also breeds the problem of the “Matthew Effect” and “resource congestion.” Under China’s urban-rural dual structure, high-quality medical resources are predominantly concentrated in prefecture-level cities and above. Counties and rural areas commonly face challenges of talent drain and aging equipment, resulting in a “reverse triangle” pattern of resource allocation [30]. At the provincial level, this spatial imbalance is reflected in the fact that for every 10% increase in urbanization rate, the gap in health resource utilization efficiency between urban and rural areas widens by 2.3% points [31].
Although existing literature has fully revealed the complexity of the impact of digital technology and its interaction with organizational and environmental factors, most studies still mainly focus on testing the net effect of a single technological factor or the interaction between two variables. The existing empirical studies on the efficiency of health resource utilization at the provincial level in China mostly rely on regression analysis methods, which are good at estimating the average net effect of variables but are difficult to capture complex causal mechanisms such as “convergence from different paths” caused by multiple concurrent conditions. Therefore, this study incorporates the digital technology system into the technology dimension of the TOE framework and explicitly constructs a configuration analysis framework for its interaction with the organizational level (economic capital) and the environmental level (spatial structure). Methodologically, it introduces fuzzy set qualitative comparative analysis (fsQCA) to break through the limitations of traditional regression methods and systematically identify multiple concurrent causal paths under the synergy of economic capital, digital technology, and spatial structure from a configuration perspective, and then constructs a configuration model of inter-provincial health resource utilization efficiency in China (as shown in Fig. 1).
Fig. 1.
Configuration model diagram of factors influencing health resource utilization efficiency
Method
Data sources and indicator selection
Taking into account the practical considerations of data availability and integrity, this study treats the 31 provinces of China as the units of analysis, excluding Hong Kong, Macao, and Taiwan. Per capita GDP, per capita disposable income, number of broadband internet users, number of mobile internet users, and urbanization rate are sourced from the China Statistical Yearbook 2022. The number of hospital beds per 1,000 people and the number of practicing physicians per 1,000 people are obtained from the 2022 China Health Statistical Yearbook. Population density is calculated as the ratio of year-end population to land area for each of the 31 provinces, based on the 2022 China Statistical Yearbook. The digital finance index is taken from the Digital Inclusive Finance Index Report compiled and published by Peking University in 2021. The selection of indicators is as follows:
Outcome variable
The outcome variable in this study is the Health Resource Utilization Efficiency. This variable is intended to measure the relatively optimal output level that the healthcare system can achieve given a certain level of resource input. It serves as a core indicator for assessing health system performance and the degree of resource allocation optimization [32]. To ensure scientific rigor and comparability, this study adopts Data Envelopment Analysis to measure provincial-level health resource utilization efficiency in China by constructing an input-output indicator system. In the selection of input indicators, the study follows the principles of comprehensiveness and data availability, covering three core dimensions: human resources, material resources, and financial resources [33]. Material resources are measured by the “number of beds in medical and health institutions per 1,000 people,” reflecting infrastructure and capital stock. Financial resources are measured by “per capita government health expenditure,” representing the extent of public health investment by the government. Human resources are measured by the “number of practicing (assistant) physicians per 1,000 people,” indicating the core labor input for healthcare services. In selecting output indicators, both service output volume and final health outcomes are considered [34].“Life expectancy per capita” is selected as a comprehensive indicator reflecting the overall health level of the population, and the “infant mortality rate”—a metric particularly sensitive to medical and health services—is used as a supplementary indicator, together representing the ultimate outcome of health production. The “annual number of outpatient visits per capita” is used as a measure of access to and intensity of use of healthcare services within the service process output. In this way, an indicator system comprising three inputs and three outputs is constructed. The resulting health resource utilization efficiency values for each province (typically relative efficiency values between 0 and 1) serve as the outcome variable in subsequent fsQCA analysis.
Condition variables
Factors influencing the efficiency of health resource utilization are mainly divided into three primary conditions: economic capital, spatial structure, and digital technology. Population distribution and urbanization level, as key indicators reflecting spatial structure characteristics of a region, significantly affect the efficiency of health resource utilization [35, 36]; hence, population density and urbanization level are selected as condition variables for spatial structure. In terms of socioeconomic factors, economic growth rate is significantly and positively correlated with medical expenditure growth; the more developed the economy, the better the government’s capacity to continually invest budget surpluses in the health sector, thereby stimulating increased medical demand and improving system efficiency [37]. At the household level, as income rises, although the proportion of medical spending may drop, there is a shift toward higher-quality medical services with improved payment capacity, which can activate latent demand and overall utilization increases significantly [38]. Therefore, per capita GDP and per capita disposable income are chosen as condition variables. Regarding digital technology, regional internet penetration rates are significantly positively associated with public health efficiency. Mobile networks and digital financial platforms are conducive to reducing healthcare transaction costs and alleviating resource misallocation, thereby “indirectly enhancing health resource utilization efficiency“ [39]. Thus, internet broadband access users, mobile internet subscribers and the digital finance index are selected as condition variables.
Data envelopment analysis
Data Envelopment Analysis is a quantitative analysis tool based on multidimensional input and output indicators. Utilizing linear programming models, it provides a relative efficiency assessment of comparable, homogeneous decision-making units [40]. DEA is widely used for analyzing the relative efficiency of input and output in various healthcare services. In this study, the DEA-BCC index model is employed to analyze the health resource utilization efficiency of 31 provinces in China. The comprehensive efficiency value generated by this model can be further decomposed into the product of pure technical efficiency and scale efficiency, essentially quantifying the linear programming relationship in systems with multiple inputs and outputs. The assessment of DEA effectiveness centers on the comprehensive efficiency value: when this value is 1, the decision unit is deemed DEA efficient, indicating that the utilization of health resources has reached an optimal state; conversely, if the value exceeds 1, it falls into the non-DEA efficient category, suggesting that there is still room to improve the input-output configuration of health resources and that utilization efficiency can be further enhanced.
Fuzzy-set qualitative comparative analysis (fsQCA)
Fuzzy-set qualitative comparative analysis (fsQCA), originally proposed by sociologist Ragin [41], is a configuration-oriented research method focusing on “causal complexity.” Its core lies in the calibration of fuzzy set variables and the use of set theory and Boolean algebra logic to analyze how combinations of multiple antecedent conditions collaboratively drive the outcome variable. fsQCA is suitable for examining complex social science issues involving multi-factor interactions [42, 43]. In this paper, the method is applied following these steps and principles: (1) Data Calibration. Calibration refers to assigning set memberships to cases, essentially assigning membership degrees to each case. The membership degree for each case is determined based on the relative position within the sample; (2) Necessity Analysis. The core of this step is to identify the necessary antecedent conditions that drive the outcome variable, generally setting the consistency threshold at 0.9 or above to meet the scientific criteria for necessary conditions [44]; (3) Analysis of Condition Configurations. This step focuses on analyzing the sufficiency effect of different combinations of antecedent conditions on the outcome variable, where consistency represents the closeness of the causal link between the antecedent condition (or its configuration) and the outcome variable, and coverage reflects the explanatory power of a specific condition configuration for the outcome variable [45].
Necessary condition analysis
This study further integrates the Necessary Condition Analysis method to address the limitation of fsQCA—namely, its reliance on qualitative judgments alone in necessity testing. NCA not only enables the identification of whether an antecedent condition constitutes a necessary condition for the outcome but also characterizes the magnitude of necessity of the condition and its required minimum threshold through quantitative metrics such as effect size (d-value) and bottleneck level. This facilitates the quantitative delineation of key bottleneck factors constraining health resource utilization efficiency.
Data measurement and calibration
Before conducting fuzzy-set qualitative comparative analysis, variables must be calibrated to determine the degree of case membership in particular sets. In this study, both the outcome variable and the antecedent conditions are converted into membership scores ranging from 0 to 1. Complete membership, crossover point, and complete non-membership thresholds are set based on the 95th, 50th, and 5th percentiles of the sample data, respectively. Calibration points for each variable are shown in Table 1.
Table 1.
Descriptive statistics and data calibration results
| Variable | Indicator Description | Full Membership (95%) | Crossover Point (50%) | Full Non membership (5%) |
|---|---|---|---|---|
| Y | Health Resource Utilization Efficiency | 1 | 0.941 | 0.792 |
| X1 | Urbanization Rate | 88.22 | 63.42 | 45.27 |
| X2 | Population Density | 1991.6 | 285.1 | 6.33 |
| X3 | GDP per Capita | 177,770 | 65,026 | 44,778 |
| X4 | Per capita disposable income | 81882.06 | 41443.8 | 34845.9 |
| X5 | Internet broadband access users | 3810.0 | 1278.1 | 147.7 |
| X6 | Mobile Internet Subscribers | 10957.3 | 3599.3 | 460.3 |
| X7 | Digital Financial Index | 450.85 | 363.61 | 335.60 |
Results
Current Status of health resource utilization efficiency in China
In 2021, the average values for China’s health resource utilization efficiency were 0.918 for comprehensive efficiency, 0.924 for pure technical efficiency, and 0.993 for scale efficiency. Among these, 10 provinces (accounting for 32.2% of the total sample, including Beijing, Tianjin, Shanghai, Zhejiang, Fujian, Jiangxi, and Guangdong, achieved DEA efficiency, while approximately two-thirds of the provinces remained inefficient. Notably, 87.1% of the provinces (27 in total) exhibited issues such as redundant resource inputs, insufficient outputs, or unreasonable scale allocation, indicating significant room for improvement. Regarding returns to scale characteristics, six provinces including Inner Mongolia and Liaoning exhibited diminishing returns to scale, indicating excessive resource inputs. Conversely, 14 provinces such as Hebei, Shanxi, Anhui, Henan, and Hubei demonstrated increasing returns to scale, suggesting insufficient resource allocation. Detailed data is presented in Table 2.
Table 2.
Data envelopment analysis efficiency results by province
| Province | Technical Efficiency (TE) | Scale Efficiency (SE) | Overall Efficiency (Jan et al.) | Returns to Scale | DEA Effectiveness |
|---|---|---|---|---|---|
| Beijing | 1.000 | 1.000 | 1.000 | Constant | DEA Strongly Effective |
| Tianjin | 1.000 | 1.000 | 1.000 | Constant | DEA Strongly Effective |
| Hebei | 0.842 | 0.985 | 0.829 | Increasing | Non-DEA Effective |
| Shanxi | 0.851 | 0.998 | 0.849 | Increasing | Non-DEA Effective |
| Inner Mongolia | 0.809 | 1.000 | 0.809 | Decreasing | Non-DEA Effective |
| Liaoning | 0.867 | 0.970 | 0.841 | Decreasing | Non-DEA Effective |
| Jilin | 0.791 | 1.000 | 0.790 | Decreasing | Non-DEA Effective |
| Heilongjiang | 0.846 | 0.993 | 0.841 | Decreasing | Non-DEA Effective |
| Shanghai | 1.000 | 1.000 | 1.000 | Constant | DEA Strongly Effective |
| Jiangsu | 0.838 | 0.990 | 0.829 | Decreasing | Non-DEA Effective |
| Zhejiang | 1.000 | 1.000 | 1.000 | Constant | DEA Strongly Effective |
| Anhui | 0.917 | 0.997 | 0.914 | Increasing | Non-DEA Effective |
| Fujian | 1.000 | 1.000 | 1.000 | Constant | DEA Strongly Effective |
| Jiangxi | 1.000 | 1.000 | 1.000 | Constant | DEA Strongly Effective |
| Shandong | 0.794 | 0.987 | 0.784 | Decreasing | Non-DEA Effective |
| Henan | 0.864 | 0.991 | 0.856 | Increasing | Non-DEA Effective |
| Hubei | 0.900 | 0.997 | 0.897 | Increasing | Non-DEA Effective |
| Hunan | 0.878 | 0.998 | 0.876 | Increasing | Non-DEA Effective |
| Guangdong | 1.000 | 1.000 | 1.000 | Constant | DEA Strongly Effective |
| Guangxi | 0.982 | 1.000 | 0.982 | Increasing | Non-DEA Effective |
| Hainan | 1.000 | 1.000 | 1.000 | Constant | DEA Strongly Effective |
| Chongqing | 1.000 | 1.000 | 1.000 | Constant | DEA Strongly Effective |
| Sichuan | 0.882 | 0.992 | 0.875 | Increasing | Non-DEA Effective |
| Guizhou | 0.942 | 0.973 | 0.917 | Increasing | Non-DEA Effective |
| Yunnan | 0.986 | 0.970 | 0.956 | Increasing | Non-DEA Effective |
| Tibet | 1.000 | 1.000 | 1.000 | Constant | DEA Strongly Effective |
| Shaanxi | 0.857 | 0.996 | 0.854 | Increasing | Non-DEA Effective |
| Gansu | 0.954 | 0.974 | 0.929 | Increasing | Non-DEA Effective |
| Qinghai | 0.919 | 0.995 | 0.915 | Increasing | Non-DEA Effective |
| Ningxia | 1.000 | 1.000 | 1.000 | Constant | DEA Strongly Effective |
| Xinjiang | 0.941 | 0.978 | 0.920 | Increasing | Non-DEA Effective |
| National Average | 0.924 | 0.993 | 0.918 | ||
Necessity analysis of conditional variables
First, the Necessary Condition Analysis [46] method was employed to conduct the necessity analysis, with the analysis implemented via the NCA package in R software. A condition is deemed a necessary condition for the outcome if two criteria are met: the effect size (d ≥0.1) and the statistical significance (P < 0.05) [47]. Additionally, the NCA package in R provides two algorithms—ceiling regression (CR) and ceiling envelope (CE)—which are applied to continuous and discrete variables, respectively. Table 3 reports the results of the NCA-based necessity analysis. The effect sizes of all seven antecedent variables were less than 0.1. Notably, the per capita disposable income variable exhibited a statistically significant necessary effect (P < 0.05), but its effect size was small (d < 0.1), thus failing to qualify as a necessary condition for health resource utilization efficiency. Therefore, none of the aforementioned conditions constitute necessary prerequisites for enhancing the level of health resource utilization efficiency.
Table 3.
Results of necessary condition analysis
| Condition Variable | Method | Accuracy | Ceiling Zone | Scope | Effect Size (d) | P-value |
|---|---|---|---|---|---|---|
| Urbanization Rate | CR | 100.0 | 0.000 | 0.205 | 0.000 | 1.000 |
| CE | 100.0 | 0.000 | 0.205 | 0.000 | 1.000 | |
| GDP per Capita | CR | 96.8 | 0.005 | 0.201 | 0.026 | 0.221 |
| CE | 100.0 | 0.007 | 0.201 | 0.034 | 0.347 | |
| Population Density | CR | 100.0 | 0.000 | 0.203 | 0.000 | 1.000 |
| CE | 100.0 | 0.000 | 0.203 | 0.000 | 1.000 | |
| Per capita disposable income | CR | 96.8 | 0.011 | 0.199 | 0.054 | 0.042 |
| CE | 100.0 | 0.016 | 0.199 | 0.079 | 0.030 | |
| Internet broadband access users | CR | 100.0 | 0.000 | 0.199 | 0.000 | 1.000 |
| CE | 100.0 | 0.000 | 0.199 | 0.000 | 1.000 | |
| Mobile Internet Subscribers | CR | 100.0 | 0.000 | 0.205 | 0.000 | 1.000 |
| CE | 100.0 | 0.000 | 0.205 | 0.000 | 1.000 | |
| Digital Financial Index | CR | 93.5 | 0.004 | 0.201 | 0.019 | 0.218 |
| CE | 100.0 | 0.005 | 0.201 | 0.025 | 0.411 |
Second, the bottleneck level refers to the minimum threshold value of antecedent conditions that must be satisfied to attain the outcome within a specified observation scope [48]. Findings from the NCA-based bottleneck level analysis in Table 4 reveal that the constraining effects of different factors are heterogeneous. Specifically, population density, per capita disposable income, and the penetration rates of the Internet and mobile Internet exhibit no bottleneck levels (NN) across all efficiency tiers. In contrast, per capita GDP, urbanization rate, and the digital finance index demonstrate explicit bottleneck constraints at distinct thresholds. For instance, to achieve high health resource utilization efficiency at the 90th percentile, the minimum thresholds to be met are: a per capita GDP level of 14.0%, an urbanization rate of 12.1%, and a digital finance index of 7.1%.
Table 4.
Results of NCA-based bottleneck level analysis
| Health Resource Utilization Efficiency | Population Density | Urbanization Rate | GDP per Capita | Per capita disposable income | Internet broadband access users | Mobile Internet Subscribers | Digital Financial Index |
|---|---|---|---|---|---|---|---|
| 0 | NN | NN | NN | NN | NN | NN | NN |
| 10 | NN | NN | NN | NN | NN | NN | NN |
| 20 | NN | NN | NN | NN | NN | NN | NN |
| 30 | NN | NN | NN | NN | NN | NN | NN |
| 40 | NN | NN | NN | 1.5 | NN | NN | NN |
| 50 | NN | NN | NN | 4.0 | NN | NN | NN |
| 60 | NN | NN | NN | 6.5 | NN | NN | 0.4 |
| 70 | NN | NN | NN | 9.0 | NN | NN | 2.6 |
| 80 | NN | NN | 4.7 | 11.5 | NN | NN | 4.8 |
| 90 | NN | NN | 12.1 | 14.0 | NN | NN | 7.1 |
| 100 | NN | NN | 19.4 | 16.5 | NN | NN | 9.3 |
Note: CR method, NN = non-essential
This study further adopts the fsQCA method to validate the findings derived from the NCA-based necessary condition analysis. A condition is deemed a necessary condition for the outcome if its consistency level exceeds 0.9 [49]. As presented in Table 5, none of the antecedent variables achieved the threshold consistency level, indicating that no single condition alone can account for the disparities in regional health resource utilization efficiency across China. This finding corroborates the characteristic that health resource utilization efficiency is subject to the synergistic influence of multiple factors, necessitating further conditional configuration analysis to explore the combined effects of distinct antecedent conditions.
Table 5.
Necessity analysis results derived from fsQCA
| Causal Condition | High Health Resource Utilization Efficiency | Not-High Health Resource Utilization Efficiency | ||
|---|---|---|---|---|
| Consistency | Coverage | Consistency | Coverage | |
| Urbanization Rate | 0.647 | 0.698 | 0.601 | 0.560 |
| ~Urbanization Rate | 0.592 | 0.632 | 0.676 | 0.623 |
| GDP per Capita | 0.634 | 0.709 | 0.560 | 0.541 |
| ~GDP per Capita | 0.589 | 0.607 | 0.699 | 0.623 |
| Population Density | 0.567 | 0.727 | 0.489 | 0.542 |
| ~Population Density | 0.642 | 0.593 | 0.754 | 0.601 |
| Per Capita Disposable Income | 0.631 | 0.758 | 0.482 | 0.499 |
| ~Per Capita Disposable Income | 0.582 | 0.565 | 0.767 | 0.643 |
| Internet broadband access users | 0.521 | 0.574 | 0.699 | 0.665 |
| ~Internet broadband access users | 0.696 | 0.728 | 0.553 | 0.499 |
| Mobile Internet Subscribers | 0.529 | 0.581 | 0.711 | 0.673 |
| ~ Mobile Internet Subscribers | 0.702 | 0.737 | 0.558 | 0.506 |
| Digital Financial Index | 0.609 | 0.717 | 0.502 | 0.510 |
| ~Digital Financial Index | 0.583 | 0.576 | 0.721 | 0.615 |
Note: “~” denotes logical NOT
Configuration analysis
A systematic examination of the synergistic mechanisms of multiple factors was conducted, with truth tables constructed for configuration analysis in accordance with the recommendations of Pappas et al [50]. Following established research methodologies, frequency thresholds were set at 5 and consistency thresholds at 0.8, ensuring PRI consistency exceeded 0.75. Intermediate solutions constituted the fundamental analytical framework, supplemented by the identification of core and auxiliary conditions for minimal solutions. The analysis yielded five valid configuration paths, each exhibiting a consistency level of ≥ 0.75. The overall consistency was 0.724, and the total coverage was 0.688. This finding indicates that the identified configurations systematically explain approximately 68.8% of healthcare resource utilization efficiency cases. The findings indicate a satisfactory theoretical fit and a substantial explanatory capacity. The specific findings are outlined in Table 6.
Table 6.
Configurational Paths of Provincial Health Resource Utilization Efficiency in China
| Dimension | Causal Condition | H1 | H2 | H3 | H4 | H5 |
|---|---|---|---|---|---|---|
| Spatial Structure | Population Density | ⊗ | ● | ● | ⊗ | ● |
| Urbanization Rate | ⊗ | ⊗ | ⊗ | • | • | |
| Economic Capital | GDP per Capita | ⊗ | ● | ● | ● | ● |
| Per Capita Disposable Income | ● | ● | ● | |||
| Digital Technology | Internet broadband access users | ⊗ | • | • | ⊗ | ⊗ |
| Mobile Internet Subscribers | • | • | ⊗ | ⊗ | ||
| Digital Financial Index | ⊗ | • | ⊗ | • | ||
| Raw Coverage | 0.453 | 0.270 | 0.360 | 0.264 | 0.359 | |
| Unique Coverage | 0.196 | 0.004 | 0.067 | 0.001 | 0.076 | |
| Unique Coverage | 0.771 | 0.777 | 0.750 | 0.791 | 0.931 | |
| Solution Coverage | 0.688 | |||||
| Solution Consistency | 0.724 | |||||
Note: ● indicates the presence of a core condition; • indicates the presence of a peripheral condition; ⊗ indicates the absence of a core condition; ⊗ indicates the absence of a peripheral condition; blank indicates the condition is irrelevant (can be either present or absent)
Robustness check
To ensure the robustness of the research conclusions, this study adopts the robustness testing methods documented in prior literature and adjusts the calibration anchor points accordingly [51]. Initially, a reanalysis was conducted with the consistency threshold elevated from 0.80 to 0.85, thereby revealing that the fundamental conditions inherent within all five configuration pathways remained constant. The total coverage was found to have decreased from 0.688 to 0.672, while the overall consistency increased from 0.724 to 0.731. All indicator fluctuations remained within acceptable ranges. Secondly, the adjustment of the calibration anchors from the 95th, 50th, and 5th percentiles to the 90th, 50th, and 10th percentiles resulted in a highly consistent configuration structure. To further verify the robustness of the efficiency measurement, this paper re-calculates the health efficiency of each province using the super-efficiency SBM model [52]and conducts a Spearman rank correlation analysis between the obtained efficiency rankings and the original BCC-DEA model rankings. The results show that the correlation coefficient between the two is 0.938 (p < 0.01), indicating a high positive correlation. This suggests that the efficiency rankings under the two methods are highly consistent, and the core conclusion of this study has good robustness.
Discussion
The overall low efficiency of health resource utilization in China is evidenced by the uneven scale of resource investment
The data envelopment analysis indicates that a mere 11 provinces (accounting for 35.5% of the sample), including Beijing, Tianjin, and Shanghai, achieved strong DEA efficiency in health resource utilization. It is evident that the majority of provinces are characterised by inefficiency, underscoring the necessity for comprehensive enhancement of operational efficiency. The fundamental cause of this phenomenon can be traced back to a synergistic imbalance between technical efficiency and scale efficiency, which can be illustrated as follows:
At the level of resource allocation, provinces exhibiting non-DEA efficiency generally demonstrate discrepancies between resource allocation and regional health needs. The phenomenon of population mobility and the heterogeneity of health demands are not adequately reflected in resource distribution, resulting in concurrent resource shortages in areas experiencing population influx and resource underutilization in areas experiencing population outflow. This underscores the pressing necessity to refine the precision and dynamic adaptability of resource allocation. From the perspective of economies of scale, the 14 provinces exhibiting increasing returns to scale (Hebei, Shanxi, Anhui, Henan, Hubei, Hunan, etc.) demonstrate a structural imbalance between supply and demand. The concentration of high-quality health resources in the region is disproportionate, while the primary-level service capacity remains inadequate, resulting in an inverted-triangle supply pattern that is unable to meet the demand for accessible and diversified health services among the local population. Conversely, six provinces experiencing diminishing returns (Inner Mongolia, Liaoning, Jiangsu, etc.) face a development bottleneck characterised by “heavy investment but light returns.” It is evident that an increase in the physical capacity of hardware has not been accompanied by a corresponding enhancement in service capacity. This has resulted in a decline in marginal benefits derived from resource inputs. The technical efficiency dimension indicates significant shortcomings in technology application and management effectiveness. This is evidenced by the inadequate integration of medical information systems, inefficient health data sharing mechanisms, and insufficient refinement of service processes. Collectively, these factors constrain further improvements in resource utilization efficiency from a technical standpoint. It is therefore evident that in order to enhance the efficiency of health resource utilization, there is a necessity for systematic advancement in three reform areas: firstly, achieving precise resource allocation through dynamic monitoring of health needs; secondly, optimising the supply structure by promoting resource decentralisation via the tiered diagnosis and treatment system; and thirdly, strengthening technological innovation and process reengineering to enhance resource operational efficiency. It is only through the mutual support and coordinated advancement of these three dimensions that a fundamental shift can be made in health resources from scale expansion to quality and efficiency gains.
Notably, DEA results indicate that several western provinces—such as Xizang and Ningxia—exhibit DEA efficiency, which deviates from the conventional expectation that eastern provinces generally demonstrate higher efficiency. This unexpected finding may be attributed to province-specific policy support, unique geographical and demographic structures, and targeted optimization of medical resources in these regions. For example, Xizang and Ningxia may have achieved high resource utilization efficiency by concentrating resource inputs in key areas, refining resource allocation strategies, or implementing innovative healthcare service models.
Accordingly, enhancing the efficiency of health resource utilization necessitates the systematic advancement of reforms across three dimensions: realizing precision in resource allocation through dynamic monitoring of health needs; driving resource decentralization via the hierarchical medical system to optimize the supply structure; and strengthening technological innovation and process reengineering to improve resource operational efficiency. Only through the mutual support and coordinated advancement of these three dimensions can a fundamental transformation of health resources from scale expansion to quality and efficiency enhancement be achieved.
The efficiency of health resource utilization in China is influenced by the synergistic effects of multiple factors
A configuration analysis reveals that the core conditions affecting provincial-level health resource utilization efficiency in China involve the coordinated allocation of digital technology, economic capital, and spatial structure.
Configuration H1: Digital-Economic-Spatial Constraint Type. This configuration is prominently observed in less developed provinces including Gansu, Ningxia, Qinghai, Xinjiang, Yunnan, and Guizhou. Its core characteristics encompass insufficient digital technology access, weak economic capital, and spatial constraints of low population density and low urbanization rate. Taking Guizhou Province as an example: despite localized breakthroughs in the big data industry, its overall internet penetration rate and digital finance development index remain significantly below the national average. Per capita GDP and fiscal revenue are limited, while mountainous terrain contributes to population dispersion and high costs of medical service accessibility. The superposition of these digital, economic, and geographical disadvantages traps the regional health system in a “low-level equilibrium trap”: resource inputs fail to generate scale effects [53], and the promotion of digital management and telemedicine applications proceeds slowly, collectively creating systemic bottlenecks for improving health resource utilization efficiency. In response to these challenges, on the one hand, it is recommended to build a regional telemedicine center based on the provincial national health information platform, configure portable intelligent diagnosis and treatment terminals for township health centers and village clinics, and give special subsidies for network fees to medical institutions in remote areas. Through “cloud deployment and terminal sinking”, the geographical barriers can be broken at a low cost, so that high-quality medical resources can reach the “last mile”. On the other hand, through the integration of central transfer payments and local supporting funds, a provincial special fund for the integrated development of “health industry” was set up to focus on supporting health industry projects with regional characteristics. Taking Guizhou as an example, we can rely on the big data industry to develop new formats such as “Chinese medicine” and smart health tourism. At the same time, we can include telemedicine services in the scope of medical insurance payment, guide patients to stay in the county for medical treatment, and help H1 provinces break through the “low-level equilibrium trap” through the two wheel drive of “technology empowerment” and “economic hematopoiesis”, forming a virtuous cycle of “health investment - industrial value-added - efficiency improvement”.
Health industry projects with regional characteristics. Taking Guizhou as an example, we can rely on the big data industry to develop new formats such as “Internet+traditional Chinese medicine” and smart health tourism. At the same time, we can include telemedicine services in the scope of medical insurance payment, guide patients to stay in the county for medical treatment, and help H1 provinces break through the “low-level equilibrium trap” through the two wheel drive of “technology empowerment” and “economic hematopoiesis”, forming a virtuous cycle of “health investment - industrial value-added - efficiency improvement”.
Configuration H2: Economic-Digital Synergy Type. This configuration encompasses provinces including Hunan, Hubei, Jiangxi, Sichuan, and Anhui. Its core logic is that a robust economic foundation and proactive digital application can form a synergistic driving force, even in the absence of a competitive urbanization rate. Take Anhui Province as an example: amid sustained growth in per capita GDP and residents’ disposable income, it has vigorously advanced the construction of “Digital Anhui” and “Smart Healthcare,” with its provincial-level universal health information platform initially realizing data interconnection. Economic capital underpins the steady expansion of health investment, while the diffusion of digital technology optimizes resource allocation and service processes—partially mitigating the scattered distribution of resources caused by uneven urbanization levels [54], thereby achieving relatively high efficiency in health resource utilization.
Configuration H3: Comprehensive Factor Equilibrium Type. This configuration integrates three key elements: spatial structure (population density presence), economic capital (sufficient per capita GDP), and digital technology (internet users, mobile internet users, digital finance index coverage). This indicates that enhancing the utilization efficiency of healthcare resources relies on the balanced allocation of multidimensional factors. Population density ensures service scale, economic capital underpins resource investment, and digital technology enables precise resource allocation and value enhancement. Provinces in this category often establish a virtuous cycle across the entire health resource chain—“input-circulation-output.” Examples include Zhejiang, Fujian, and parts of eastern Jiangsu—regions characterized by population concentration, economic development, and deep integration of digital technology into healthcare. From primary healthcare informatization to regional health big data platform construction, these areas demonstrate relatively comprehensive systems and high resource utilization efficiency.
Configuration H4: Urbanization-Driven Model. The core prerequisites for this pathway lie in a relatively high urbanization rate and sufficient economic capital, while the application of digital technologies lags behind. Representative provinces include Liaoning and Jilin. Taking Liaoning Province as an example—once a traditional industrial base—its urbanization rate has long exceeded the national average. Urban medical resources, particularly large tertiary hospitals, are equipped with complete hardware facilities, and the province boasts a solid historical accumulation of economic strength. However, the popularization of digital medical applications and the level of data integration remain relatively inadequate, with the “information silo” phenomenon among medical institutions still prominent. This results in a health system that is highly reliant on the traditional spatial agglomeration model centered on large hospitals. It has failed to fully leverage digital tools to achieve efficient cross-institutional resource circulation and process reengineering, thereby limiting further improvements in efficiency.
Configuration H5: All-Factor Empowerment Model. This pathway embodies the optimal all-dimensional synergy among spatial structure (high population density and high urbanization rate), economic capital, and digital technologies (notably digital finance). Exemplary cases include Beijing, Shanghai, and Guangdong. Beijing epitomizes this pattern: as a megacity, its highly concentrated population and economy endow health resources with extreme scale effects and market capacity; robust public finance and residents’ payment capacity underpin the introduction of cutting-edge medical technologies and facility construction; meanwhile, Beijing serves not only as a digital technology innovation hub but also leverages tools like digital finance to advance medical insurance payment reform and health industry financing. Such deep all-factor synergy enables the health system to achieve holistic optimization spanning precise resource allocation, intelligent operation, and value realization, attaining the nation’s leading resource utilization efficiency and verifying Hollands [55] theory that “factor synergy generates superlinear returns.”
Conclusion
Based on the five differentiated paths identified via configuration analysis, this study proposes a precision policy system that transcends one-size-fits-all solutions. This system advocates that policy design must be tightly coupled with the factor endowments of individual provinces, with the implementation of “path-specific” interventions. For provinces under Path H1 (e.g., Gansu, Guizhou) trapped in the multiple constraints of “digital-economy-space”, the core lies in executing a comprehensive plan centered on external empowerment and foundational enhancement. Through targeted initiatives such as the “Digital Health New Infrastructure” special project, “resource package”-based assistance, and innovative service models for remote areas, efforts are directed at breaking the “low-level equilibrium trap”.For provinces under Paths H2 and H4 (e.g., Hunan, Anhui, Liaoning)—which possess single- or dual-dimensional advantages but face distinct shortcomings—the policy focus is on strengthening factor synergy and precisely addressing deficits. For Path H2 provinces, they should be incentivized to translate economic advantages into investments in deep digital scenarios (e.g., medical artificial intelligence) and dismantle data-sharing barriers. For Path H4 provinces, guidance is needed to leverage medical resources accumulated through urbanization, prioritizing the digital transformation of traditional medical centers to unlock the potential of existing resources.For provinces under Paths H3 and H5 (e.g., Zhejiang, Beijing)—which have achieved factor equilibrium or all-factor empowerment—policies should position them as front-runners in innovation leadership and standardization. They should be granted “pilot rights” in areas such as intelligent medical insurance cost control and health industry ecosystem construction, and tasked with developing replicable toolkits of mature “modern health governance” experiences for national scaling. To ensure the implementation of these differentiated policies, cross-regional factor flow and compensation mechanisms must be established in parallel, and efficiency improvements should be integrated into performance evaluation systems. This will form a systematic support framework that progresses from “addressing deficits” to “forging competitive strengths”.
This study transcends the traditional application of the TOE framework—which is typically used to analyze the adoption of a single technology—by developing it into an integrated analytical tool for explaining how the “technology-organization-environment” (TOE) triad of elements synergistically interacts to shape the performance of complex systems. This advancement deepens the understanding of mechanisms underlying organizational effectiveness under multiple constraints. Concurrently, the study systematically identifies five differentiated configurational paths driving the efficiency of health resource utilization across Chinese provinces, constructing a novel provincial typology rooted in “factor endowment structure”. Beyond verifying the pervasive existence of equifinality and causal asymmetry, it precisely delineates a complete efficiency spectrum spanning from “low-level equilibrium traps” to “all-factor synergy”. Nevertheless, the study has certain limitations. In terms of data: while the use of the DEA model and authoritative macro-statistical data ensures measurement comparability, efficiency indicators focus on input-output ratios and thus fail to fully capture qualitative dimensions such as service quality and patient satisfaction. Methodologically: fsQCA excels at identifying concurrent configurations and equivalent paths of multiple antecedents, yet its set-theory-based static comparative logic hinders the revelation of dynamic evolution and path dependence in inter-element synergies. Additionally, provincial-level analysis, while effective for uncovering inter-provincial disparities, may obscure significant intra-provincial heterogeneity (e.g., urban-rural or prefectural-level differences). The current provincial-level statistics mainly report macro health indicators such as average life expectancy and infant mortality rate, but lack health outcome data by disease type and population group. For instance, indicators such as the prevalence rate, control rate, and treatment cost of major chronic diseases like hypertension and diabetes are only published in some years of the National Health Services Survey, and cannot form a continuous panel. Future research could attempt to utilize medical insurance big data to construct health output indicators by disease type. These limitations illuminate directions for future research, such as employing longitudinal tracking or mixed-methods approaches to test configurational mechanisms at a more micro level, thereby achieving a more comprehensive, dynamic, and mechanistically insightful understanding.
Acknowledgements
Not applicable.
Author contributions
RanRan Diao: the acquisition, analysis, interpretation of data; the creation of new software used in the work; have drafted the work or substantively revised it.
Funding
No funding.
Data availability
The dataset generated and analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
The ethical approval committee of Nanchang University approved this study and confirmed that the study has no side effects on the participants of the study. All of the procedures were performed in accordance with the Declaration of Helsinki and relevant policies in China. Informed consent to participate was obtained from all the participants.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Bouckley T, Peiris D, Nambiar D, Mishra S, Sood T, Purwar P, Elshaug AG, Landon BE, Pearson SA, Huckel Schneider C, Schierhout G. Addressing health equity during design and implementation of health system reform initiatives: a scoping review and framework. Int J Equity Health. 2025;24(1):68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Li H, Wang R, Wang H. Regional disparities and influencing factors of health production efficiency: evidence from China. BMC Public Health. 2025;25(1):2406. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Drazin R. The processes of technological innovation. J Technol Transf. 1991;16(1):45–6. [Google Scholar]
- 4.Grossman M. On the Concept of Health Capital and the Demand for Health. J Polit Econ. 1972;80:223–55. [Google Scholar]
- 5.Squires DA. Explaining high health care spending in the United States: an international comparison of supply, utilization, prices, and quality. Issue Brief (Commonw Fund). 2012;10:1–14. [PubMed] [Google Scholar]
- 6.Bloom DE, Chen S, Kuhn M, McGovern ME, Oxley L, Prettner K. The economic burden of chronic diseases: Estimates and projections for China, Japan, and South Korea. J Econ Ageing. 2020;17:100163. [Google Scholar]
- 7.Shen Q, Chang B, Yin G, Wang W. The Impact of Health Investment on Economic Growth: Evidence from China. Iran J Public Health. 2020;49(4):684–92. [PMC free article] [PubMed] [Google Scholar]
- 8.Zhou J, Li J, Mao N, Li S, Chen X, Wang D, Zhang Y, Shi X, Gao X, Tao S, Pu X, Lu J, Hu X. The equity of health service utilization in less developed areas of China: evidence from Gansu Province. BMC Public Health. 2025;25(1):3081. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Meessen B. The Role of Digital Strategies in Financing Health Care for Universal Health Coverage in Low- and Middle-Income Countries. Glob Health Sci Pract. 2018;6(Suppl 1):S29–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Shi Y. Digital finance and corporate breakthrough innovation: Evidence from China. PLoS ONE. 2024;19(7):e0307737. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Lyu Y, Peng Y, Liu H, Hwang JJ. Impact of digital economy on the provision efficiency for public health services: empirical study of 31 provinces in China. Int J Environ Res Public Health. 2022;19(10). [DOI] [PMC free article] [PubMed]
- 12.Zhao S, Wang Y, Chen Y, Zhou M. Healthcare resource allocation and patient choice: evidence from rural China. Int J Equity Health. 2025;24(1):87. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Wei Z, Wei K, Yang J, Zhang M, Feng Y. Can the digital economy foster advancements in the healthcare sector? — a case study using interprovincial data from China. BMC Public Health. 2025;25(1):196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Yang X, Chen M, Du J, Wang Z. The inequality of inpatient care net benefit under integration of urban-rural medical insurance systems in China. Int J Equity Health. 2018;17(1):173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Shao Y, Yang X, Chen Q, Guo H, Duan X, Xu X, Yue J, Zhang Z, Zhao S, Zhang S. Determinants of digital health literacy among older adult patients with chronic diseases: a qualitative study. Front Public Health. 2025;13:1568043. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Xu X, Zhang Y, Wang Y, Zhao C, Zhang Y, Xie X. Impact of regional digital transformation on public health: an empirical analysis based on 31 provinces in China. BMC Public Health. 2025;25(1):2485. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Bogale TN, Teklehaimanot SM, Fufa Debela T, Enyew DB, Nigusse Bedada A, Dufera Kebebew S, Nigusie Weldeab A, Wolde Daka D, Willems HJ, Bekele TA. Barriers, facilitators and motivators of electronic community health information system use among health workers in Ethiopia. Front Digit Health. 2023;5. [DOI] [PMC free article] [PubMed]
- 18.Tseng CH, Chen RJ, Tsai SY, Wu TR, Tsaur WJ, Chiu HW, Yang CY, Lo YS. Exploring the COVID-19 Pandemic as a Catalyst for Behavior Change Among Patient Health Record App Users in Taiwan: Development and Usability Study. J Med Internet Res. 2022;24(1):e33399. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Ye M, Hu J. From Disconnection to Well-Being: A Longitudinal Study on Digital Access as a Social Determinant of Health for Older Adults in China. The Journals of Gerontology: Series B; 2025. [DOI] [PubMed] [Google Scholar]
- 20.Chen J, Wang S, Han Y, Zhang Y, Li Y, Zhang B, Li X, Zhang J. Geodetector analysis of individual and joint impacts of natural and human factors on maternal and child health at the provincial scale. Sci Rep. 2024;14(1):1643. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Guo L, Chai Y, Yang C, Zhang L, Guo H, Yang H. Has smart city transition elevated the provision of healthcare services? Evidence from China’s. Smart City Pilot Policy Digit Health. 2023;9:20552076231197335. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Mahmoud AO, Nkanga D, Onakoya A. Views of West African surgeons on how well their educational and professional backgrounds may have prepared them for health leadership roles. N Am J Med Sci. 2010;2(12):580–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Song N, Xu L, Chen X, Xu H, Jiang S. Research on the improvement path of grassroots social governance innovation performance in China——Qualitative comparative analysis based on 35 cases. PLoS ONE. 2024;19(2):e0297587. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Hawes SM, O’Reilly KM, Mascitelli TM, Winczewski J, Dazzio R, Arrington A, Morris KN. Measuring the One Health impacts associated with creating access to veterinary care before and during the COVID-19 pandemic. Front Public Health. 2024;12. [DOI] [PMC free article] [PubMed]
- 25.Goddard M, Smith P. Equity of access to health care services:: Theory and evidence from the UK. Soc Sci Med. 2001;53(9):1149–62. [DOI] [PubMed] [Google Scholar]
- 26.Chai Y, Yuan X, Guo L, Chen Z. The impact of broadband infrastructure construction on medical resource mismatch: quasi-natural experiment from the Broadband China Policy. J Med Internet Res. 2024;26. [DOI] [PMC free article] [PubMed]
- 27.Jiang TB, Deng ZW, Zhi YP, Cheng H, Gao Q. The Effect of Urbanization on Population Health: Evidence From China. Front Public Health. 2021;9:706982. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Chen Y, Li Q, Zhang Y, Zhang X, Zheng H. Investigating the efficiency of medical and health resources and its influencing factors in the Yangtze River Delta urban agglomerations: based on the undesirable super-efficiency SBM-Malmquist-Tobit model. Front Public Health. 2025;13. [DOI] [PMC free article] [PubMed]
- 29.Yang Q, Sun Z, Zhang H. Assessment of Urban Green Development Efficiency Based on Three-Stage DEA: A Case Study from China’s. Yangtze River Delta Sustain. 2022;14(19):12076. [Google Scholar]
- 30.Hong Z, Li N, Li D, Li J, Li B, Xiong W, Lu L, Li W, Zhou D. Telemedicine During the COVID-19 Pandemic: Experiences From Western China. J Med Internet Res. 2020;22(5):e19577. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Zhang L, Liu Y. Urbanization and the Urban-Rural Efficiency Gap in Health Resources: Evidence from 31 Chinese Provinces. Chin J Health Policy. 2022;15(4):301–12. [Google Scholar]
- 32.Jacobs R, Smith P, Street A. Measuring efficiency in health care: analytic techniques and health policy. Cambridge;: Cambridge University Press; 2006. [Google Scholar]
- 33.Hollingsworth B. Non-Parametric and Parametric Applications Measuring Efficiency in Health Care. Health Care Manag Sci. 2003;6(4):203–18. [DOI] [PubMed] [Google Scholar]
- 34.O’Neill L, Rauner M, Heidenberger K, Kraus M. A cross-national comparison and taxonomy of DEA-based hospital efficiency studies. Socio-Economic Plann Sci. 2008;42(3):158–89. [Google Scholar]
- 35.Qiu L, Yang L, Li H, Wang L. The impact of health resource enhancement and its spatiotemporal relationship with population health. Front Public Health. 2022;10:1043184. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Xu A, Wei H. Does the fairness of healthcare resource allocation affect utilization efficiency?—An empirical study based on China’s provincial panel data. Front Health Serv. 2025;5. [DOI] [PMC free article] [PubMed]
- 37.Kwilinski A, Vysochyna A. Efficiency of Healthcare Financing: Case of European Countries. Int J Financial Stud. 2024;12(3):87. [Google Scholar]
- 38.Gong J, Shi L, Wang X, Sun G. The efficiency of health resource allocation and its influencing factors: evidence from the super efficiency slack based model-Tobit model. Int Health. 2023;15(3):326–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Zhao X, Wu S, Yan B, Liu B. New evidence on the real role of digital economy in influencing public health efficiency. Sci Rep. 2024;14(1):7190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Sahil MA, Lohani QMD. Comprehensive intuitionistic fuzzy network data envelopment analysis incorporating undesirable outputs and shared resources. MethodsX. 2024;12:102710. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Ragin CC. The comparative method. Moving beyond qualitative and quantitative strategies. University of California Press. 2014.
- 42.Liao TF. Fuzzy-Set Social Science. By Charles C. Ragin. University of Chicago Press, 2000. 352 pp. Cloth, $48.00; paper, $18.95. Soc Forces. 2001;80(1):354–6.
- 43.Chew X, Gatea Atiyah A, Alnoor A, Abbas S, Muhsen YR, Bayram GE, Alnoor A, et al. Cham, Springer Nature Switzerland: 111–124.
- 44.Schneider CQ, Wagemann C. Set-Theoretic Methods for the Social Sciences: A Guide to Qualitative Comparative Analysis. Cambridge: Cambridge University Press; 2012. [Google Scholar]
- 45.Llopis-Albert C, Palacios-Marqués D, Simón-Moya V. Fuzzy set qualitative comparative analysis (fsQCA) applied to the adaptation of the automobile industry to meet the emission standards of climate change policies via the deployment of electric vehicles (EVs). Technol Forecast Soc Chang. 2021;169:120843. [Google Scholar]
- 46.Houpy JC, Lee WW, Woodruff JN, Pincavage AT. Medical student resilience and stressful clinical events during clinical training. Med Educ Online. 2017;22(1):1320187. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Jan D, Erwin VDL, Roelof K. A statistical significance test for necessary condition analysis. Organizational Res Methods. 2018;21:109442811879527.
- 48.Dul J. Necessary condition analysis (NCA): Logic and methodology of \necessary but not sufficient\ causality. Organizational Res Methods. 2016;19(1):10–52. [Google Scholar]
- 49.Ragin CC, Fiss P. Net effects analysis versus configurational analysis: An empirical demonstration. Redesigning Social Inquiry: Fuzzy Sets beyond. 2008;190:212. [Google Scholar]
- 50.Kahneman D, Tversky A. Prospect theory: An analysis of decision under risk. New York, NY, US: Cambridge University Press; 1988. [Google Scholar]
- 51.Cangialosi N. Fuzzy-Set Qualitative Comparative Analysis (fsQCA) in Organizational Psychology: Theoretical Overview, Research Guidelines, and A Step-By-Step Tutorial Using R Software. Span J Psychol. 2023;26:e21. [DOI] [PubMed] [Google Scholar]
- 52.Tone K. A slacks-based measure of super-efficiency in data envelopment analysis. Eur J Oper Res. 2002;143(1):32–41. [Google Scholar]
- 53.Hu Z. How digital economy and green and low carbon policies affect non-agricultural employment?—Evidence from China. Humanit Social Sci Commun. 2025;12(1):1260. [Google Scholar]
- 54.Cuadros DF, Huang Q, Mathenjwa T, Gareta D, Devi C, Musuka G. Unlocking the potential of telehealth in Africa for HIV: opportunities, challenges, and pathways to equitable healthcare delivery. Front Digit Health. 2024;6. [DOI] [PMC free article] [PubMed]
- 55.Hollands RG. Will the real smart city please stand. up? City. 2008;12(3):303–20. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The dataset generated and analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.