Spatio-temporal distribution characteristics of COVID-19 in China: a city-level modeling study

Qianqian Ma; Jinghong Gao; Wenjie Zhang; Linlin Wang; Mingyuan Li; Jinming Shi; Yunkai Zhai; Dongxu Sun; Lin Wang; Baozhan Chen; Shuai Jiang; Jie Zhao

doi:10.1186/s12879-021-06515-8

Spatio-temporal distribution characteristics of COVID-19 in China: a city-level modeling study

BMC Infect Dis. 2021 Aug 14;21(1):816. doi: 10.1186/s12879-021-06515-8.

Authors

Qianqian Ma^#^{1

2}, Jinghong Gao^#^{1

2}, Wenjie Zhang^{1

2}, Linlin Wang^{1

2}, Mingyuan Li^{1

2}, Jinming Shi^{1

2}, Yunkai Zhai^{1

2

3}, Dongxu Sun^{1

2}, Lin Wang^{1

2}, Baozhan Chen^{1

2}, Shuai Jiang^{1

2}, Jie Zhao^{4

5}

Affiliations

¹ The First Affiliated Hospital of Zhengzhou University, Zhengzhou, Henan, China.
² National Engineering Laboratory for Internet Medical Systems and Applications, Zhengzhou, China.
³ School of Management Engineering, Zhengzhou University, Zhengzhou, China.
⁴ The First Affiliated Hospital of Zhengzhou University, Zhengzhou, Henan, China. zhaojie@zzu.edu.cn.
⁵ National Engineering Laboratory for Internet Medical Systems and Applications, Zhengzhou, China. zhaojie@zzu.edu.cn.

^# Contributed equally.

Abstract

Background: The coronavirus disease 2019 (COVID-19) has become a pandemic. Few studies have been conducted to investigate the spatio-temporal distribution of COVID-19 on nationwide city-level in China.

Objective: To analyze and visualize the spatiotemporal distribution characteristics and clustering pattern of COVID-19 cases from 362 cities of 31 provinces, municipalities and autonomous regions in mainland China.

Methods: A spatiotemporal statistical analysis of COVID-19 cases was carried out by collecting the confirmed COVID-19 cases in mainland China from January 10, 2020 to October 5, 2020. Methods including statistical charts, hotspot analysis, spatial autocorrelation, and Poisson space-time scan statistic were conducted.

Results: The high incidence stage of China's COVID-19 epidemic was from January 17 to February 9, 2020 with daily increase rate greater than 7.5%. The hot spot analysis suggested that the cities including Wuhan, Huangshi, Ezhou, Xiaogan, Jingzhou, Huanggang, Xianning, and Xiantao, were the hot spots with statistical significance. Spatial autocorrelation analysis indicated a moderately correlated pattern of spatial clustering of COVID-19 cases across China in the early phase, with Moran's I statistic reaching maximum value on January 31, at 0.235 (Z = 12.344, P = 0.001), but the spatial correlation gradually decreased later and showed a discrete trend to a random distribution. Considering both space and time, 19 statistically significant clusters were identified. 63.16% of the clusters occurred from January to February. Larger clusters were located in central and southern China. The most likely cluster (RR = 845.01, P < 0.01) included 6 cities in Hubei province with Wuhan as the centre. Overall, the clusters with larger coverage were in the early stage of the epidemic, while it changed to only gather in a specific city in the later period. The pattern and scope of clusters changed and reduced over time in China.

Conclusions: Spatio-temporal cluster detection plays a vital role in the exploration of epidemic evolution and early warning of disease outbreaks and recurrences. This study can provide scientific reference for the allocation of medical resources and monitoring potential rebound of the COVID-19 epidemic in China.

Keywords: COVID-19; China; Geographic hotspot; Spatio-temporal distribution; Time–space scan; Visualization.

MeSH terms

COVID-19*
China / epidemiology
Cities / epidemiology
Humans
Pandemics
SARS-CoV-2
Spatio-Temporal Analysis