Spatiotemporal dynamics and driving factors of human resources for health in traditional Chinese medicine in China

Abstract

The uneven distribution of human resources for health (HRH) in China, particularly within the realm of traditional Chinese medicine (TCM), has long posed a significant challenge. Although prior studies have examined regional disparities in overall HRH, limited research has specifically addressed the spatial and temporal dynamics of HRH in TCM and their underlying determinants. This study employs a multiscale geographically weighted regression (MGWR) model to explore the spatial and temporal dynamics of HRH in TCM across 31 Chinese provinces from 2008 to 2021. MGWR allows each explanatory variable to operate at its optimal spatial scale, capturing localized patterns of spatial heterogeneity. The results reveal that although HRH in TCM has generally increased over time, substantial disparities persist, with eastern and central regions exhibiting a higher degree of resource agglomeration while western regions continue to lag behind. Economic factors such as wage income, the number of TCM institutions, and education funding emerge as the most significant and spatially heterogeneous determinants. These localized effects suggest that region-specific policy interventions are needed-such as financial incentives and infrastructure support for underserved western regions, and institutional integration and quality standardization in eastern areas-reflecting the differentiated roles and development contexts of TCM across regions. By leveraging the advantages of MGWR in capturing multiscale spatial patterns, this study provides empirical evidence to support more targeted and sustainable planning for HRH in TCM development. Enhancing the accessibility of TCM services not only addresses current regional imbalances but also strengthens the healthcare system's capacity to respond to the needs of an aging population and future public health emergencies. The findings offer valuable insights for policymakers aiming to improve health equity in China and may inform similar efforts in other countries facing spatial disparities in traditional medicine resources.

Keywords: Human resources for health; Regional disparities spatiotemporal analysis; Traditional Chinese medicine.

Fig. 1

Location of the study area. This map was generated by ArcGIS 22.0 (https://www.esri.com/).

Fig. 2

Overview of this study.

Fig. 3

Changes in agglomeration degree of HRH in each province from 2008 to 2021.

Fig. 4

Variation trend of TCM practitioners per 1000 people in each province from 2008 to 2021.

Fig. 5

Changes in Geographical of HRH in TCM in each province from 2008 to 2021.

Fig. 6

Comparison of HRH results obtained using different models.

Fig. 7

Spatial distribution maps of agglomeration degree of HRH between 2008 and 2021. Panels A to D illustrate the agglomeration degree in different periods, including 2008 (A), 2008–2013 (B), 2014–2018 (C), and 2020–2021 (D). Darker green shading indicates a higher degree of agglomeration, with value ranges classified by natural breaks. Provinces with no available data are marked with gray hatching. Maps were generated by ArcGIS 22.0 (https://www.esri.com/).

Fig. 8

Moran’s I scatter plot of HRH in TCM across Chinese provinces in 2008 (A), 2018 (B), and 2021 (C). The positive slopes and Moran’s I values indicate significant and increasing spatial clustering over time.

Fig. 9

Kernel density spatial distribution of HRH in TCM across China in 2008 (A), 2018 (B), and 2021 (C). High-density clusters are consistently observed in eastern and central regions. Maps were generated by ArcGIS 22.0 (https://www.esri.com/).

Fig. 10

Spatial distribution in the standard deviation ellipse of HRH in China in 2008 and 2021. This map was generated by ArcGIS 22.0 (https://www.esri.com/).

Fig. 11

Spatial distribution of the regression coefficients of the influential factors in the MGWR model in 2008: (A) spatial distribution of regression coefficients for unemployment rate; (B) spatial distribution of regression coefficients for average wage; (C) spatial distribution of regression coefficients for education funding; (D) spatial distribution of regression coefficients for number of TCM institution; (E) spatial distribution of regression coefficients for per-capita healthcare expenditure. Maps were generated by ArcGIS 22.0 (https://www.esri.com/).

Fig. 12

Spatial distribution of the regression coefficients of the influential factors in the MGWR model from 2009 to 2013: (A) spatial distribution of regression coefficients for number of TCM institutions; (B) spatial distribution of regression coefficients for per-capita healthcare expenditure. Maps were generated by ArcGIS 22.0 (https://www.esri.com/).

Fig. 13

Spatial distribution of the regression coefficients of the influential factors in the MGWR model from 2014 to 2018: (A) spatial distribution of regression coefficients for average wage of healthcare personnel; (B) spatial distribution of regression coefficients for education funding; (C) spatial distribution of regression coefficients for number of TCM institutions; (D) spatial distribution of regression coefficients for per-capita healthcare expenditure. Maps were generated by ArcGIS 22.0 (https://www.esri.com/).

Fig. 14

Spatial distribution of the regression coefficients of the influential factors in the MGWR model from 2019 to 2021: (A) spatial distribution of regression coefficients for average wage of healthcare personnel; (B) spatial distribution of regression coefficients for education funding; (C) spatial distribution of regression coefficients for number of TCM institutions; (D) spatial distribution of regression coefficients for per-capita healthcare expenditure. Maps were generated by ArcGIS 22.0 (https://www.esri.com/).