Gao R, Yu S C, Wang Q Q, Zhou X H, Liu N K, Tan F
Chinese Center for Disease Control and Prevention, Beijing 102206, China.
Peking University Health Science Center, Beijing 100191, China.
Zhonghua Liu Xing Bing Xue Za Zhi. 2022 Mar 10;43(3):297-304. doi: 10.3760/cma.j.cn112338-20211217-00996.
Based on the geographic information systems, we exploreed the spatiotemporal clustering and the development and evolution of COVID-19 epidemic at prefectural level in China from the time when the epidemic was discovered to the time when the lockdown ended in Wuhan. The information and data of the confirmed COVID-19 cases from December 8, 2019 to April 8, 2020 were collected from 367 prefectures in China for a spatial autocorrelation analysis with software GeoDa, and software ArcGIS was used to visualize the results. Software SatScan was used for spatiotemporal scanning analysis to visualize the hot-spot areas of the epidemic. The incidence of new cases of COVID-19 had obvious global autocorrelation and the partial autocorrelation results showed that incidence of COVID-19 had different spatial distribution at different times from December 8, 2019 to March 4, 2020. There was no significant difference in global autocorrelation coefficient from March 5, 2020 to April 8, 2020. The statistical analysis of spatiotemporal scanning identified two kinds of spatiotemporal clustering areas, the first class clustering areas included 10 prefectures, mainly distributed in Hubei, from January 13 to February 25, 2020. The secondary class clustering areas included 142 prefectures, mainly distributed in provinces in the north and east of Hubei, from January 23 to February 1, 2020. There was a clear spatiotemporal correlation in the distribution of the outbreaks in the early phase of COVID-19 epidemic (December 8, 2019-March 4, 2020) in China. With the decrease of the case and effective prevention and control measures, the epidemics had no longer significant correlations among areas from March 5 to April 8. The study results showed relationship with time points of start and adjustment of emergency response at different degree in provinces. Furthermore, improving the early detection of new outbreaks and taking timely and effective prevention and control measures played an important role in blocking the transmission.
基于地理信息系统，我们探究了中国地级行政区从新冠肺炎疫情发现到武汉解封期间的时空聚集性以及疫情的发展演变。收集了2019年12月8日至2020年4月8日中国367个地级行政区新冠肺炎确诊病例的信息和数据，使用GeoDa软件进行空间自相关分析，并运用ArcGIS软件对结果进行可视化。使用SatScan软件进行时空扫描分析以可视化疫情热点地区。新冠肺炎新增病例数具有明显的全局自相关性，局部自相关结果显示，2019年12月8日至2020年3月4日不同时间新冠肺炎发病率具有不同的空间分布。2020年3月5日至4月8日全局自相关系数无显著差异。时空扫描统计分析识别出两类时空聚集区，第一类聚集区包括10个地级行政区，主要分布在湖北，时间为2020年1月13日至2月25日。第二类聚集区包括142个地级行政区,主要分布在湖北北部和东部省份，时间为2020年1月23日至2月1日。中国新冠肺炎疫情早期（2019年12月8日至2020年3月4日）疫情分布存在明显的时空相关性。随着病例数下降和有效防控措施实施，3月5日至4月8日各地区间疫情不再具有显著相关性。研究结果显示省级应急响应启动和调整时间点在不同程度上与疫情相关。此外,提高新发疫情早期发现能力并及时采取有效防控措施对阻断传播起到重要作用。
