Regional differences in the level of healthcare in China Authors: Hongrui Cui, Qingli Tan, Xiaoshan Liu Published: 2026 ABSTRACT ObjectivesLarge inequities exist between different regions in China in terms of access to healthcare. However, there is limited quantitative research on these regional disparities. Therefore, we aimed to analyse the differences in the levels of healthcare in various regions of China, and to provide a reference for healthcare reform aimed at optimising the allocation of medical resources in China.MethodsWe collected and organised healthcare-related data from China’s provinces, autonomous regions, and municipalities directly under the Central Government from 2012 to 2021 and selected 18 indicators related to medical services for the year 2021. We used factor analysis, cluster analysis, and the Theil and Atkinson index to quantitatively study the medical service levels of China’s regions.ResultsThere was a large discrepancy between regions in terms of the level of healthcare, with the lowest level of healthcare in economically underdeveloped regions. The distribution of medical resources in various regions of China is uneven, with a lack of medical resources in the western and central regions and underutilisation of medical resources in easten region.ConclusionsTo rectify the imbalance in the level of healthcare services across China, health reform must focus on enhancing regional economic development. Strengthen the capacity of grassroots healthcare services and reduce inequalities in medical resource allocation. Moreover, future policies should focus more on addressing regional disparities in healthcare funding and urban-rural gaps in medical resource allocation. Introduction Access to healthcare is essential for both health and well-being. Beyond being a practical concern for individuals, it can significantly influence a country’s economy [1–4]. In recent times, the discrepancy between the supply and demand of healthcare services has attracted considerable attention at the national level in China. On August 26, 2016, President Xi Jinping chaired a meeting of the Political Bureau of the Communist Party of China Central Committee, where the outline of the “Healthy China 2030” plan was deliberated and approved.This marked the first time since the founding of the People’s Republic of China that a medium- and long-term strategic plan was formulated at the national level for the health sector. It ushered in a new stage of development in promoting a healthy China, bringing both opportunities and challenges. In recent years, China has implemented various measures to improve medical services across regions. However, these efforts have yet to meet the population’s health needs, and significant disparities persist between regions. To support health reform policies, a comprehensive evaluation of medical services across China’s 31 provinces, autonomous regions, and municipalities is urgently needed, using objective and detailed indicators. In this study, we analyzed 2021 indicators of medical services using factor analysis, cluster analysis, and the Theil index to assess the level of medical services across China’s provinces, autonomous regions, and municipalities. Previous studies have comprehensively evaluated the levels of medical services in China [5]. However, existing studies usually select a variety of indicators, such as the number of beds and medical personnel per 1,000 people, to construct an evaluation system, and then use the entropy method to measure the level of development of health services [6, 7]. Other scholars have focused on the efficiency dimension of medical service quality and used the data envelopment analysis model to evaluate the efficiency of healthcare services [8, 9]. A large number of studies have revealed different degrees of spatial differences and convergence characteristics in China in terms of public health expenditures, medical supply capacity, and medical allocation efficiency [10–12]. However, the current research has limitations. First, although there are many published studies on regional disparities in medical and health services in China [13], most of the data comes from urban or rural studies [14–16]. In addition, many nationwide analyses are province-based, and concentrate mainly on the micro-analysis of a certain province as well as urban or rural areas [17, 18], and very few province-to-province research findings are based on the national level [19]. Second, most of the analyses are concerned with the unilateral assessment of the development status of the regional medical and health services [20, 21], and research on the impact of the economy on health services is limited. Third, most existing studies adopted the same method; namely, the spatial entropy weighting method [22, 23]. Therefore, there is limited visual and quantitative research on regional disparities and development conditions. In the current study, we analysed data from the National Statistical Yearbook 2022, China Health Statistics Yearbook 2022, and official websites to select 18 relevant indicators. Using factor analysis, cluster analysis, and the Theil index, we aimed to compare the differences in healthcare services in different regions of China and provide a theoretical basis and rationale for the reform of healthcare services in China. Methods Data sources and processing Based on the differentiation approach of the 2022 China Health Statistics Yearbook, this study divided China into four parts: western, central, eastern, and northeastern. The western provinces and regions include Xinjiang, Ningxia, Qinghai, Gansu, Shanxi, Tibet, Yunnan, Guizhou, Sichuan, Chongqing, and Guangxi. The central region includes Hunan, Hubei, Henan, Jiangxi, Anhui, and Inner Mongolia along with seven parts of Shaanxi. The eastern region includes Hainan, Guangdong, Shandong, Fujian, Zhejiang, Jiangsu, Shanghai, Hebei, Tianjin, and parts of Beijing. The northeastern region includes Heilongjiang, Jilin, and Liaoning. This study does not include data from Taiwan, Macau, or Hong Kong. The data for this study were derived from the China Health Statistics Yearbook 2022 and the China Statistics Yearbook 2022. Guided by the principles of indicator feasibility and data availability, this study constructed a healthcare service evaluation system through the Delphi method to consolidate expert consensus. Drawing upon established evaluation frameworks [5, 24–26] and grounded in the theoretical foundations of equality of opportunity, process equity, and outcome equity, the developed indicator system is systematically presented in Table 1. Table 1Indicator system for assessment of the level of regional medical servicesIndicatorNotationNumber of hospital beds per 1,000 population \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{1}$$\end{document} Number of medical personnel per 1,000 population \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{2}$$\end{document} Total number of outpatient visits to medical institutions \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{3}$$\end{document} Average outpatient medical cost per visit \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{4}$$\end{document} Average number of physician diagnoses and treatments per day \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{5}$$\end{document} Daily inpatient beds per doctor \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{6}$$\end{document} Average length of stay (days) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{7}$$\end{document} Bed occupancy rate (%) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{8}$$\end{document} Life expectancy in 2020 (years) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{9}$$\end{document} Per capita disposable income (yuan) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{10}$$\end{document} Tertiary industry in local GDP (%) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{11}$$\end{document} Total health costs in 2020 (hundred million yuan) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{12}$$\end{document} Per capita expenditure on healthcare (yuan) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{13}$$\end{document} Emergency mortality (%) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{14}$$\end{document} Number of live births (number) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{15}$$\end{document} Number of medical and health institutions (number) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{16}$$\end{document} Number of health technicians per thousand population \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{17}$$\end{document} Hospital mortality (%) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{18}$$\end{document} Table 1Indicator system for assessment of the level of regional medical servicesIndicatorNotationNumber of hospital beds per 1,000 population \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{1}$$\end{document} Number of medical personnel per 1,000 population \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{2}$$\end{document} Total number of outpatient visits to medical institutions \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{3}$$\end{document} Average outpatient medical cost per visit \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{4}$$\end{document} Average number of physician diagnoses and treatments per day \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{5}$$\end{document} Daily inpatient beds per doctor \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{6}$$\end{document} Average length of stay (days) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{7}$$\end{document} Bed occupancy rate (%) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{8}$$\end{document} Life expectancy in 2020 (years) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{9}$$\end{document} Per capita disposable income (yuan) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{10}$$\end{document} Tertiary industry in local GDP (%) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{11}$$\end{document} Total health costs in 2020 (hundred million yuan) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{12}$$\end{document} Per capita expenditure on healthcare (yuan) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{13}$$\end{document} Emergency mortality (%) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{14}$$\end{document} Number of live births (number) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{15}$$\end{document} Number of medical and health institutions (number) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{16}$$\end{document} Number of health technicians per thousand population \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{17}$$\end{document} Hospital mortality (%) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{18}$$\end{document} Indicator system for assessment of the level of regional medical services Research methodology Factor analysis Data preprocessing Because different indicators have different scales and are not comparable, raw data must be standardised to eliminate the effect of the scale. The standardisation formula was as follows:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{Z}_{ij}=\frac{{X}_{ij}-{X}_{min}}{{X}_{max}-{X}_{min}}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{ij}$$\end{document} represents the raw data for the jth sample of indicator i, which is the evaluation index of the medical service level in each region, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{Z}_{ij}$$\end{document} is the standardised data. Kaiser–Meyer–Olkin (KMO) and Bartlett’s test Before conducting the factor analysis, the variables were subjected to Bartlett’s test to determine the p-values and KMO values.The KMO value was 0.662(> 0.6), which is suitable for factor analysis. The results of Bartlett’s test showed p < 0.001, indicating that the selected indicators were suitable for factor analysis. Factor extraction and interpretation The eigenvalues and variance contributions of the factors were obtained using SPSS 22.0 software (IBM Corp., Armonk, NY, USA) (Table 2). Table 2Total variance contribution of factor analysis method for 2021 dataFactorInitial eigenvalueRotational sums of squared loadingsTotalVariance (%)Cumulative (%)TotalVariance (%)Cumulative (%) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{1}$$\end{document} 6.45835.87735.8776.41935.66235.662 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{2}$$\end{document} 4.48124.89660.7733.72120.67356.335 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{3}$$\end{document} 2.40913.38274.1542.57414.29870.633 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{4}$$\end{document} 1.6989.43683.5902.33212.95783.590 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{5}$$\end{document} 0.8924.95388.543 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{6}$$\end{document} 0.6243.46592.008 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{7}$$\end{document} 0.4212.33894.347 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{8}$$\end{document} 0.2741.52395.870 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{9}$$\end{document} 0.2491.38497.254 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{10}$$\end{document} 0.1440.79998.054 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{11}$$\end{document} 0.1170.64898.702 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{12}$$\end{document} 0.0790.44099.142 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{13}$$\end{document} 0.0520.29299.434 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{14}$$\end{document} 0.0360.19899.631 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{15}$$\end{document} 0.0280.15399.785 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{16}$$\end{document} 0.0180.10299.887 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{17}$$\end{document} 0.0120.06799.954 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\:{X}_{18}$$\end{document} 0.0080.046100.000 Table 2Total variance contribution of factor analysis method for 2021 dataFactorInitial eigenvalueRotational sums of squared loadingsTotalVariance (%)Cumulative (%)TotalVariance (%)Cumulative (%) \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{1}$$\end{document} 6.45835.87735.8776.41935.66235.662 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{2}$$\end{document} 4.48124.89660.7733.72120.67356.335 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{3}$$\end{document} 2.40913.38274.1542.57414.29870.633 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{4}$$\end{document} 1.6989.43683.5902.33212.95783.590 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{5}$$\end{document} 0.8924.95388.543 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{6}$$\end{document} 0.6243.46592.008 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{7}$$\end{document} 0.4212.33894.347 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{8}$$\end{document} 0.2741.52395.870 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{9}$$\end{document} 0.2491.38497.254 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{10}$$\end{document} 0.1440.79998.054 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{11}$$\end{document} 0.1170.64898.702 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{12}$$\end{document} 0.0790.44099.142 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{13}$$\end{document} 0.0520.29299.434 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{14}$$\end{document} 0.0360.19899.631 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{15}$$\end{document} 0.0280.15399.785 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{16}$$\end{document} 0.0180.10299.887 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{X}_{17}$$\end{document} 0.0120.06799.954 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\:{X}_{18}$$\end{document} 0.0080.046100.000 Total variance contribution of factor analysis method for 2021 data Table 2 shows that the first four metrics explained 83.590% of the total variance, indicating that the four extracted metrics represented 83.590% of the original six indicators that measured the level of medical services. This indicates minimal information loss, allowing the four metrics Y1, Y2, Y3, and Y4 to better explain the initial data. Y1 demonstrated significant loading on multiple factors, including per capita disposable income, Tertiary industry in local GDP (%), average outpatient medical cost per visit, per capita expenditure on health care, life expectancy in 2020, number of medical personnel per 1,000 population, number of health technicians per thousand population, average number of physician diagnoses and treatments per day, and hospital mortality (Table 3). These factors were strongly correlated and grouped into one category, which mainly reflects healthcare investment in medical personnel and medical cost resources, as well as residents’ health levels in each province and city. Y2 demonstrated large loadings on the total number of outpatient visits to medical institutions, total health costs in 2020, number of live births, and number of medical and health institutions, which mainly reflects the investment in medical equipment resources and medical costs. Y3 had large loadings on bed occupancy rate and daily inpatient beds per doctor, which can be grouped together to reflect the efficiency of medical resources. Y4 demonstrated large loadings on emergency mortality, number of hospital beds per 1,000 people, and average length of hospital stay (LOS), which can be grouped together to reflect the quality of the resources of medical equipment and outpatient services. Table 3Rotated factor loading matrix for 2021FactorY1Y2Y3Y4Per capita disposable income (yuan)0.9240.1180.158-0.191Average outpatient medical cost per visit0.906Per capita expenditure on healthcare (yuan)0.9010.271Tertiary industry in local GDP (%)0.8420.175-0.174Life expectancy in 2020 (years)0.8280.277Number of medical personnels per 1,000 population0.783-0.477Number of health technicians per 1,000 population0.774-0.197-0.229Average number of physician diagnoses and treatments per day0.6590.1280.545-0.274Hospital mortality (%)0.649-0.129-0.2960.438Total number of outpatient visits to medical institutions0.1130.9410.224Total health costs in 2020 (hundred million yuan)0.2310.9250.157Number of live births (number)-0.1610.9110.219-0.124Number of medical and health institutions (number)-0.2290.8750.238Bed occupancy rate (%)0.2490.3810.794Daily inpatient beds per doctor-0.3380.1810.7730.469Emergency mortality(%)0.200-0.6830.405Number of hospital beds per 1,000 population-0.257-0.1150.893Average length of hospital stay (days)0.2030.1260.805Factor rotation was performed using the maximum variance method Table 3Rotated factor loading matrix for 2021FactorY1Y2Y3Y4Per capita disposable income (yuan)0.9240.1180.158-0.191Average outpatient medical cost per visit0.906Per capita expenditure on healthcare (yuan)0.9010.271Tertiary industry in local GDP (%)0.8420.175-0.174Life expectancy in 2020 (years)0.8280.277Number of medical personnels per 1,000 population0.783-0.477Number of health technicians per 1,000 population0.774-0.197-0.229Average number of physician diagnoses and treatments per day0.6590.1280.545-0.274Hospital mortality (%)0.649-0.129-0.2960.438Total number of outpatient visits to medical institutions0.1130.9410.224Total health costs in 2020 (hundred million yuan)0.2310.9250.157Number of live births (number)-0.1610.9110.219-0.124Number of medical and health institutions (number)-0.2290.8750.238Bed occupancy rate (%)0.2490.3810.794Daily inpatient beds per doctor-0.3380.1810.7730.469Emergency mortality(%)0.200-0.6830.405Number of hospital beds per 1,000 population-0.257-0.1150.893Average length of hospital stay (days)0.2030.1260.805Factor rotation was performed using the maximum variance method Rotated factor loading matrix for 2021 Factor rotation was performed using the maximum variance method Factor score The matrix of the component score coefficients was calculated (Table 4). Table 4Matrix of component score coefficients for 2021coefficient1234Number of hospital beds per 1,000 population-0.028-0.0610.1040.394Number of health technicians per 1,000 population0.118-0.039-0.0510.024Total number of outpatient visits to medical institutions0.0110.265-0.039-0.040Average outpatient medical cost per visit0.141-0.012-0.008-0.020Average number of physician diagnoses and treatments per day0.117-0.0470.241-0.084Daily inpatient beds per doctor-0.028-0.0630.3520.243Average length of hospital stay0.0350.038-0.0100.346Bed occupancy rate (%)0.059-0.0040.3270.068Life expectancy in 2020 (years)0.1300.0660.0170.039Per capita disposable income (yuan)0.1470.0040.074-0.067Tertiary industry in local GDP (%)0.137-0.0470.104-0.057Total health costs in 2020 (billions of dollars)0.0290.267-0.060-0.006Per capita expenditure on healthcare (yuan)0.143-0.0140.0130.123Emergency mortality rate (%)0.0150.069-0.2800.140Number of live births (number)-0.0320.260-0.047-0.058Number of medical and health institutions (number)-0.0480.286-0.1540.084Number of medical personnel per 1,000 population0.1100.041-0.190-0.016Hospital mortality (%)0.099-0.012-0.0740.182 \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned}{Y}_{1}&=-0.028{X}_{1}+0.118{X}_{2}+0.011{X}_{3}\cr&\quad+0.141{X}_{4}+0.117{X}_{5}-0.028{X}_{6}\cr&\quad+0.035{X}_{7}+0.059{X}_{8}+0.130{X}_{9}\cr&\quad+0.147{X}_{10}+0.137{X}_{11}+0.029{X}_{12}\cr&\quad+0.143{X}_{13}+0.015{X}_{14}-0.032{X}_{15}\cr&\quad-0.048{X}_{16}+0.110{X}_{17}+0.099{X}_{18}\end{aligned}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned}{Y}_{2}&=-0.061{X}_{1}-0.039{X}_{2}+0.265{X}_{3}\cr&\quad-0.012{X}_{4}-0.047{X}_{5}-0.063{X}_{6}\cr&\quad+0.038{X}_{7}-0.004{X}_{8}+0.066{X}_{9}\cr&\quad+0.004{X}_{10}-0.047{X}_{11}-0.267{X}_{12}\cr&\quad-0.014{X}_{13}+0.069{X}_{14}+0.260{X}_{15}\cr&\quad+0.286{X}_{16}+0.041{X}_{17}-0.012{X}_{18}\end{aligned}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned}{Y}_{3}&=0.104{X}_{1}-0.051{X}_{2}-0.039{X}_{3}\cr&\quad-0.008{X}_{4}+0.241{X}_{5}+0.352{X}_{6}\cr&\quad-0.010{X}_{7}+0.327{X}_{8}+0.017{X}_{9}\cr&\quad+0.074{X}_{10}-0.104{X}_{11}-0.060{X}_{12}\cr&\quad+0.013{X}_{13}-0.280{X}_{14}-0.047{X}_{15}\cr&\quad-0.154{X}_{16}-0.190{X}_{17}-0.074{X}_{18}\end{aligned}$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned}{Y}_{4}&=0.394{X}_{1}+0.024{X}_{2}-0.040{X}_{3}\cr&\quad-0.020{X}_{4}-0.084{X}_{5}+0.243{X}_{6}\cr&\quad+0.346{X}_{7}+0.068{X}_{8}+0.039{X}_{9}\cr&\quad-0.067{X}_{10}-0.057{X}_{11}-0.006{X}_{12}\cr&\quad+0.123{X}_{13}+0.140{X}_{14}-0.058{X}_{15}\cr&\quad+0.084{X}_{16}-0.016{X}_{17}+182{X}_{18}\end{aligned}$$\end{document} Table 4Matrix of component score coefficients for 2021coefficient1234Number of hospital beds per 1,000 population-0.028-0.0610.1040.394Number of health technicians per 1,000 population0.118-0.039-0.0510.024Total number of outpatient visits to medical institutions0.0110.265-0.039-0.040Average outpatient medical cost per visit0.141-0.012-0.008-0.020Average number of physician diagnoses and treatments per day0.117-0.0470.241-0.084Daily inpatient beds per doctor-0.028-0.0630.3520.243Average length of hospital stay0.0350.038-0.0100.346Bed occupancy rate (%)0.059-0.0040.3270.068Life expectancy in 2020 (years)0.1300.0660.0170.039Per capita disposable income (yuan)0.1470.0040.074-0.067Tertiary industry in local GDP (%)0.137-0.0470.104-0.057Total health costs in 2020 (billions of dollars)0.0290.267-0.060-0.006Per capita expenditure on healthcare (yuan)0.143-0.0140.0130.123Emergency mortality rate (%)0.0150.069-0.2800.140Number of live births (number)-0.0320.260-0.047-0.058Number of medical and health institutions (number)-0.0480.286-0.1540.084Number of medical personnel per 1,000 population0.1100.041-0.190-0.016Hospital mortality (%)0.099-0.012-0.0740.182 Matrix of component score coefficients for 2021 Finally, we substituted the standardised raw data into the formula to obtain Y1, Y2, Y3, and Y4. To accurately analyse the level of healthcare services in the 31 provinces, cities, and autonomous regions, a weighting process was required to obtain the composite score model:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned}Y&=\frac{0.35662}{0.83590}{Y}_{1}+\frac{0.20673}{0.83590}{Y}_{2}\cr&\quad+\frac{0.14298}{0.83590}{Y}_{3}+\frac{0.129573}{0.83590}{Y}_{4}\end{aligned}$$\end{document} Cluster analysis Existing studies utilizing Moran’s I or Getis-Ord indices typically assess spatial autocorrelation based on a single measurable variable. In contrast, this study measures healthcare levels through a comprehensive evaluation system comprising 18 indicators. The k-means cluster analysis was selected as it accommodates multidimensional input variables, enabling the integration of multiple evaluation criteria. Methodologically, this represents a robust and innovative approach that aligns with the multidimensional nature of our assessment framework. SPSS software was used for cluster analysis of the data in Table 5, with the number of clusters ranging from 2 to 5. The final clustering results are presented in Table 6. Table 5Comprehensive scores for the level of medical services in 31 provinces, cities, and autonomous regions of China in 2021ProvinceY1Y2Y3Y4YComprehensive rankingBeijing3.63-0.61-1.05-0.631.122Tianjin1.38-0.31-0.73-1.480.168Hebei-0.330.88-1.59-0.21-0.2323Shanxi-0.480.56-1.290.66-0.1820Inner Mongolia-0.180.27-1.80.43-0.2525Liaoning0.29-0.09-0.931.450.177Jilin0.260.15-1.231.370.1510Heilongjiang-0.07-0.29-1.462.1-0.0315Shanghai2.73-0.922.120.721.411Jiangsu0.44-0.260.62-0.340.186Zhejiang0.880.120.68-1.220.335Anhui-0.42-0.110.38-0.32-0.1921Fujian-0.29-0.010.69-1.29-0.2122Jiangxi-0.71-0.041.07-0.09-0.1418Shandong0.140.66-0.79-0.180.0612Henan-0.520.660.260.350.0413Hubei-0.17-0.340.870.520.0711Hunan-0.31-0.111.070.860.168Guangdong0.260.460.51-1.940.0114Guangxi-0.46-0.021.18-0.4-0.0616Hainan-0.15-0.330.21-0.77-0.2323Chongqing0.14-0.371.30.980.344Sichuan-0.060.261.041.550.463Guizhou-0.9-0.241.13-0.02-0.2525Yunnan-0.73-0.281.36-0.1-0.1619Tibet-1.460.8-1.61-1.82-0.9831Shanxi-0.24-0.04-0.020.37-0.0616Gansu-0.880.03-0.03-0.07-0.3827Qinghai-0.670.05-0.940.3-0.3928Ningxia-0.44-0.24-0.52-0.95-0.4830Xinjiang-0.69-0.27-0.520.17-0.4329 Table 5Comprehensive scores for the level of medical services in 31 provinces, cities, and autonomous regions of China in 2021ProvinceY1Y2Y3Y4YComprehensive rankingBeijing3.63-0.61-1.05-0.631.122Tianjin1.38-0.31-0.73-1.480.168Hebei-0.330.88-1.59-0.21-0.2323Shanxi-0.480.56-1.290.66-0.1820Inner Mongolia-0.180.27-1.80.43-0.2525Liaoning0.29-0.09-0.931.450.177Jilin0.260.15-1.231.370.1510Heilongjiang-0.07-0.29-1.462.1-0.0315Shanghai2.73-0.922.120.721.411Jiangsu0.44-0.260.62-0.340.186Zhejiang0.880.120.68-1.220.335Anhui-0.42-0.110.38-0.32-0.1921Fujian-0.29-0.010.69-1.29-0.2122Jiangxi-0.71-0.041.07-0.09-0.1418Shandong0.140.66-0.79-0.180.0612Henan-0.520.660.260.350.0413Hubei-0.17-0.340.870.520.0711Hunan-0.31-0.111.070.860.168Guangdong0.260.460.51-1.940.0114Guangxi-0.46-0.021.18-0.4-0.0616Hainan-0.15-0.330.21-0.77-0.2323Chongqing0.14-0.371.30.980.344Sichuan-0.060.261.041.550.463Guizhou-0.9-0.241.13-0.02-0.2525Yunnan-0.73-0.281.36-0.1-0.1619Tibet-1.460.8-1.61-1.82-0.9831Shanxi-0.24-0.04-0.020.37-0.0616Gansu-0.880.03-0.03-0.07-0.3827Qinghai-0.670.05-0.940.3-0.3928Ningxia-0.44-0.24-0.52-0.95-0.4830Xinjiang-0.69-0.27-0.520.17-0.4329 Comprehensive scores for the level of medical services in 31 provinces, cities, and autonomous regions of China in 2021 Table 6Final clustering resultscategory 1Beijing, Shanghaicategory 2Zhejiang, Chongqing, Sichuancategory 3Tianjin, Liaoning, Jilin, Heilongjiang, Jiangsu, Shandong, Henan, Hubei, Hunan, Guangdong, Guangxi, Shaanxicategory 4Hebei, Shanxi, Inner Mongolia, Anhui, Fujian, Jiangxi, Hainan, Guizhou, Yunnan, Gansu, Qinghai, Ningxia, Xinjiangcategory 5Tibet Table 6Final clustering resultscategory 1Beijing, Shanghaicategory 2Zhejiang, Chongqing, Sichuancategory 3Tianjin, Liaoning, Jilin, Heilongjiang, Jiangsu, Shandong, Henan, Hubei, Hunan, Guangdong, Guangxi, Shaanxicategory 4Hebei, Shanxi, Inner Mongolia, Anhui, Fujian, Jiangxi, Hainan, Guizhou, Yunnan, Gansu, Qinghai, Ningxia, Xinjiangcategory 5Tibet Final clustering results The Theil Index and the Atkinson Index The Theil index, a key measure of income disparity between individuals and regions, was used to analyse the intra- and inter-regional differences in healthcare services across China, along with national differences and contribution rates.1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:T=\frac{1}{k}\sum\nolimits_{q=1}^{k}(\frac{{Y}_{q}}{\overline{S}}\times\:\mathrm{l}\mathrm{n}\frac{{Y}_{q}}{\overline{S}})$$\end{document}2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{T}_{p}=\frac{1}{{k}_{p}}\sum\nolimits_{q=1}^{{k}_{p}}(\frac{{Y}_{pq}}{\overline{{Y}_{p}}}\times\:\mathrm{l}\mathrm{n}\frac{{Y}_{pq}}{\overline{Y}})$$\end{document}3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned}T&={T}_{w}+{T}_{b}\cr&=\sum\nolimits_{p=1}^{4}(\frac{{k}_{p}}{k}\times\:\frac{{\overline{Y}}_{p}}{\overline{Y}}\times\:{T}_{p})\cr&\quad+\sum\nolimits_{p=1}^{4}\left(\frac{{k}_{p}}{k}\times\:\frac{\overline{Y}}{\overline{Y}}\times\:\mathrm{ln}\frac{\overline{{Y}_{p}}}{\overline{Y}}\right)\end{aligned}$$\end{document} In Eq. (1), T represents the overall difference in the level of healthcare services in China’s regional Theil index, the size of which is in [0,1]. The smaller the Theil index, the smaller the overall difference in the level of healthcare service development, and vice versa, indicating that the overall difference is greater. q denotes the province, k denotes the number of provinces,\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\:Y}_{q}$$\end{document} denotes the level of medical services in province q, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\overline{Y}$$\end{document} denotes the average value of the healthcare service level in each region of China. In Eq. (2), \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{T}_{p}$$\end{document} denotes the overall Theil index of variation for region p, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{k}_{p}$$\end{document} denotes the number of provinces in region p, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{Y}_{pq}$$\end{document} denotes the level of healthcare service development in province q of region p, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\overline{Y}}_{p}$$\end{document} denotes the average value of healthcare service development in region p. Equation (3) further decomposes the Theil index of the overall variation in the level of medical services into the Theil index of intra-regional variation \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{T}_{w}$$\end{document} and the inter-regional variation Theil index, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{T}_{b}$$\end{document}. In addition, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{T}_{w}$$\end{document} / T and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{T}_{b}$$\end{document} / T are defined as the contributions of intra- and inter-regional variations to the overall variation. The Atkinson index was originally an index that measures the notion of social welfare norms in income distribution unfairness. It is mainly used to measure the equity of income and has gradually become adopted in the field of health resources and healthcare service [27]. The index is given by the following formula:\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{A}_{\varepsilon\:}=1-{\left(\frac{1}{N}{\sum\nolimits_{i=1}^{N}\left(\frac{{Y}_{i}}{\mu\:}\right)}^{1-\varepsilon\:}\right)}^{\frac{1}{1-\varepsilon\:}}\text{}(\varepsilon\:\ne\:1)$$\end{document} \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{A}_{{\upepsilon\:}}$$\end{document} is the Atkinson index; i is province i; \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{Y}_{i}$$\end{document} is Comprehensive scores (Table 5); N is and ε is used to assume the level of inequality aversion, which reflects the strength of society’s aversion to inequality. When the value of ε rises, relatively more weight is attached to transfers at the lower end of the distribution and relatively less weight is attached at the upper end. The coefficient ε in the present study was set at 0.5 to calculate the Atkinson index [27]. The Atkinson index ranges from 0 to1, the more equal the distribution, the value of it is more close to 1. Results Factor analysis results Factor analysis scores and rankings showed that the level of healthcare in each region was closely related to the level of economic development. Beijing and Shanghai, which are both areas of high economic development, showed substantial advantages in terms of the level of healthcare. Beijing’s factor 1 score reached 3.63, reflecting abundant medical resources. However, its Factor 3 and Factor 4 scores of -1.05 and − 0.63 indicate underutilisation of these resources. Jiangsu and Zhejiang demonstrated similar results to that of Beijing. Shanghai, as a national economic centre, had a high factor 1 score of 2.73. Unlike Beijing, Shanghai’s factor 3 and 4 scores were also very high, at 2.12 and 0.72, respectively, and the total score was the highest, indicating that Shanghai makes appropriate use of its healthcare resources and provides better healthcare services. Chongqing and Sichuan had high scores on factors 3 and 4, whereas Sichuan had a relatively low score on factor 1, indicating a shortage of medical resources. Jiangxi, Guizhou, Yunnan, Tibet, Gansu, Qinghai, and Xinjiang were the seven provinces and autonomous regions with factor 1 scores lower than − 0.67, indicating a relative shortage of medical resources, with Tibet showing the lowest factor 1 score of only − 1.46. Heilongjiang had the highest factor 4 score of 2.1, indicating that its medical equipment resources and outpatient service quality are superior. Jiangxi, Hunan, Guangxi, Sichuan, Chongqing, Guizhou, and Yunnan, and the seven provinces and autonomous regions had factor 3 scores higher than 1.00, indicating that the utilisation of their medical resources is more efficient. Cluster analysis results and analysis The spatial distribution of healthcare levels in various regions of China is shown in Fig. 1. The figure demonstrates superior medical service levels in Shanghai and Beijing and inferior services in Tibet. Overall, the distribution of medical resources is unbalanced. Medical service levels are relatively low in regions such as Hebei, Shanxi, Inner Mongolia, Anhui, Jiangxi, Guizhou, and others, with particularly unsatisfactory conditions in economically underdeveloped areas of central and western China. Fig. 1Spatial distribution of the level of healthcare in China in 2021. Note: Based on the standard map with review number GS (2019) 1822 downloaded from the standard map service website of the Map Technology Review Center of the Ministry of Natural Resources, with no modifications to the base map Spatial distribution of the level of healthcare in China in 2021. Note: Based on the standard map with review number GS (2019) 1822 downloaded from the standard map service website of the Map Technology Review Center of the Ministry of Natural Resources, with no modifications to the base map Analysis of the Theil index and Atkinson index measurement Since the Y values in Table 5 include negative numbers, the Theil index cannot be directly computed. To address this while preserving the relative disparities between variables, this study adopts a methodological adjustment by adding 1 to all original values, ensuring all transformed values are strictly positive [28]. The Theil index of the level of healthcare in various regions in 2021 is shown in Table 7. The overall Theil index was 0.546, indicating large differences in the overall level of healthcare in various regions of China. The decomposition results show that the contribution of intra-regional disparity reached 74.68% in 2021, which is greater than 50%, indicating marked inequities in the level of healthcare services in different places. While the contribution rate of inter-regional disparity was only 25.32%, the contribution rate of intra-regional disparity was much larger than that of inter-regional disparity, indicating that the overall inequality in the level of healthcare in various regions of China are mainly attributed to intra-regional differences.Further decomposition results of the intra-regional disparity in 2021 showed that the Theil index of healthcare level in the eastern, central, western, and northeastern regions of China was 0.586, 0.156, 0.011, and 0.002. Regional healthcare resource allocation exhibits a significant positive correlation with economic development levels. The Atkinson Index analysis reveals significant disparities in healthcare resource allocation across regions. As shown in Table 6, the western region demonstrates the highest level of healthcare inequality (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{A}_{0.5}=0.092$$\end{document}), followed by the eastern region (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{A}_{0.5}=0.039$$\end{document}). This pattern is partially attributable to the unique composition of western China, where two major metropolitan centres (Chengdu and Chongqing) are classified as Category 2 resource-abundant areas through cluster analysis, while other western provinces predominantly fall into resource-constrained Category 3. Notably, although the eastern region possesses the most advanced economy and abundant medical resources overall, it exhibits pronounced intra-regional imbalances. In contrast, the northeastern region, despite receiving less healthcare investment than eastern and central China, demonstrates the most equitable distribution pattern (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{A}_{0.5}=0.002$$\end{document}), suggesting more effective resource allocation mechanisms may exist in this area. These findings underscore the spatial complexity of healthcare equity that cannot be fully explained by conventional economic development indicators alone. Table 7Regional differences in the level of medical services in the East, Central, West, and Northeast based on the Theil indexEasternCentralWestenNortheastenIntraregional disparityInterregional disparityContribution to intraregional disparity (%)Contribution of interregional disparity (%)Overall inequalityTheil Index0.5860.1560.0110.0020.4230.14374.67925.3210.546Atkinson Index0.0390.0060.0920.002////0.055 Table 7Regional differences in the level of medical services in the East, Central, West, and Northeast based on the Theil indexEasternCentralWestenNortheastenIntraregional disparityInterregional disparityContribution to intraregional disparity (%)Contribution of interregional disparity (%)Overall inequalityTheil Index0.5860.1560.0110.0020.4230.14374.67925.3210.546Atkinson Index0.0390.0060.0920.002////0.055 Regional differences in the level of medical services in the East, Central, West, and Northeast based on the Theil index Discussion Enhancing the balance of inter-regional and intra-region economic development By fully leveraging the collaborative efforts of the government, enterprises, markets, and social forces to meet diverse medical security needs, it is essential to sustain and deepen strategic initiatives such as the “Rise of Central China,” “Western Development,” “Pairing Assistance,” and “Northeast Revitalization.” These measures will promote economic development in central, western, and northeastern regions, solidify the material foundation for advancing healthcare services in these areas, and ultimately narrow regional disparities in medical security levels. For example, by offering financial subsidies [29], tax concessions [30], and investments in healthcare infrastructure construction. Appropriately increasing financial investment in healthcare in easten economically disadvantaged subregion encourages social and personal investment in healthcare and jointly raising the level of development of health services [21]. Second, with the goal of promoting equality in basic healthcare, the fiscal transfer system [31] should be further improved by gradually reducing tax refunds through progressive reforms and increasing general and balanced subsidies to economically underdeveloped regions in the area of public health. Special subsidies can be used to provide vocational and technical training for grassroots medical and nursing staff to improve their professional standards. When the financial resources for health care in a given area are insufficient, consideration should be given to the provision of special financial subsidies for public health by the three levels of government; namely, the city, province, and central government. Strengthen the capacity of grassroots healthcare services and reduce inequalities in medical resource allocation In the densely populated central and eastern regions, the demand for medical security is relatively high, yet inequitable distribution remains more pronounced. In the northeastern region, with a more severe aging population, demand for basic public health services remains high despite exhibiting the lowest levels of inequality. In the vast and sparsely populated western regions, where accessibility to medical resources is limited in some areas, the critical importance of grassroots healthcare systems is self-evident. Promoting service capacity building in local medical institutions will continue to enhance the grassroots capacity for disease prevention, treatment, and health management, steadily expand the scale of free medical student training in rural areas, and further increase the proportion of Community Health Service Centers and township health centres that meet national capacity standards. This will promote continuous improvement in the quality of medical services, the availability of facilities for elderly patients [32], and modification of hospital wards [33, 34]. The recruitment of professional nursing personnel will be expanded [35, 36], thereby continuously improving the level of medical and nursing services. These critical initiatives aim to eliminate imbalanced urban-rural service distribution, enhance equity and accessibility of medical services for rural populations, and progressively narrow the gap in medical security levels between urban and rural areas [19]. However, the Theil and Atkinson index values presented in Table 7 reveal significantly higher healthcare disparities in both eastern and western regions. These inequities are fundamentally attributable to the entrenched urban-rural divide in basic public health service equalization, with observed disparities exhibiting progressive escalation. Policy responses should therefore focus on rebalancing resource distribution toward rural primary care facilities and establishing retention incentives to counteract urban-centric concentration of medical workforce. Improving the healthcare insurance system to narrow regional and urban-rural disparities Improving the healthcare insurance system is one of the key means to narrow regional healthcare service disparities and urban-rural gaps, in addition to fiscal transfer payments through national subsidies and the introduction of private capital. The transition from the “New Rural Cooperative Medical Scheme” to the establishment of a “Unified Basic Medical Insurance System for Urban and Rural Residents” marks a pivotal step in China’s efforts to reduce disparities in healthcare resource allocation. However, under the “dual-track” socioeconomic system between urban and rural areas, significant disparities in healthcare coverage remain evident. Future reforms in healthcare insurance need to first rely on national-level coordination to achieve policy unification of basic healthcare security systems across regions, ensuring residents enjoy essentially consistent basic healthcare service levels. Second, reforms should guide residents’ healthcare-seeking behaviors to prevent excessive demand for high-quality care from disproportionately flowing to tertiary hospitals in developed cities. Third, it is necessary to appropriately expand the gradient differences in tiered healthcare reimbursement, strengthen the diversionary effect of pricing mechanisms on patients, and further advance the reform of the tiered diagnosis and treatment system. This study has several limitations. A positive correlation is observed between the level of healthcare services and regional economic development [37]. Regional economic development in China exhibits discernible spillover effects, which may contribute to spatial autocorrelation characteristics in healthcare service levels. However, the methodologies employed in this study, including the Theil index and cluster analysis, primarily serve to delineate the current status of China’s healthcare provision. These approaches do not explicitly address the spatial autocorrelation patterns or their underlying mechanisms. Future research will aim to explore the intricate relationships between regional economic development, healthcare service quality, and resource allocation patterns through spatially explicit analytical frameworks. Conclusion Since China started its new medical reforms in 2009, the distribution status of workforce have been continuously optimized. However, there are still some problems. Imbalanced resource distribution persists across China’s regions. While the economically developed eastern regions demonstrate stronger healthcare service provision capacity, disparities remain most pronounced in these areas. In contrast, most western and central regions grapple with insufficient supply alongside persistent inequities. In the short term, adopting technological interventions such as telemedicine models can help alleviate healthcare service disparities. Additionally, significant urban-rural disparities persist and warrant targeted policy attention. But in the future, the policies should clearly addressed the problems in allocation of health resources and try to narrow the gap in regional financial investment and health expenditure.