Zhongguo quanke yixue (Dec 2023)
Evaluation of Medical Level in China by Provinces Based on Principal Component Analysis and TOPSIS Model
Abstract
Background During the nationwide epidemic of COVID-19 infection, the spatial agglomeration of medical resources in China has been highlighted, and there are obvious differences in medical level among provinces. Currently, the evaluation of medical level in China by provinces was mainly conducted by domestic scholars using quantitative methods, while comprehensive method was less applied to evaluate the medical level by provinces. Objective To understand the differences in the level of healthcare development in China by provinces, so as to provide a reference for healthcare decision makers. Methods In November 2022, CNKI, Wanfang Data Knowledge Service Platform, and Web of Science were searched by computer for the researches in the field of medical level. Based on the existing research results, relative and average indicators were selected to construct the evaluation index system. The data of each evaluation index was extracted or calculated by using China Health and Health Statistical Yearbook 2022 as the data source. Using the principal component analysis and TOPSIS model, the medical levels of 31 provinces in China (Hong Kong Special Administrative Region, Macao Special Administrative Region and Taiwan Province were not included in the statistics) were comprehensively evaluated. Results A total of 6 qualified papers were retrieved and 13 relative and average indicators were selected from three aspects of medical resources, medical services, and medical security to construct the evaluation system. The KMO value was 0.733, and Bartlett's spherical test showed that χ2=346.908, P<0.001, suggesting that the data were suitable for principal component analysis; four principal components were extracted according to the criterion of characteristic root above 1.000, including the scale of medical resources and quality of medical services (F1), the efficiency of medical institutions (F2), infectious disease control ability (F3), and other factors (F4), and the cumulative percent variance of the four principal components was 84.012%. After establishing the linear model of each principal component based on the matrix of the principal component scores, the comprehensive evaluation model for the medical level was obtained based on the cumulative percent variance of the four principal components: Y=0.439 85×Y1+0.158 54×Y2+0.154 40×Y3+0.087 34×Y4. The top three provinces in terms of comprehensive score of medical level were Beijing (151.908 points), Shanghai (124.379 points), and Tianjin (78.673 points). The TOPSIS proximity ranking showed that Beijing and Shanghai were at the top level (proximity was 0.767 and 0.646, respectively), and the 31 provinces could be divided into three echelons with proximity 0.400 and 0.201 as the nodes. The first echelon included three provinces of Beijing, Shanghai and Tianjin, the second echelon included 25 provinces such as Zhejiang Province and Sichuan Province, the third echelon included three provinces of Hebei Province, Ningxia Hui Autonomous Region and Tibet Autonomous Region. Conclusion There is an obvious imbalance in the level of medical development in China by provinces, showing an olive-shaped structure of "big in the middle and small at the two ends" in the overall distribution of medical level in 31 provinces. The government should increase the incline degree of policy for provinces with low ranking in medical level, such as Hebei Province, play a coordinating role in regional health planning, and implement targeted assistance by using telemedicine and medical big data.
Keywords
