Early characteristics of the COVID-19 outbreak predict the subsequent epidemic size

Zhang, Lei; Tao, Yusha; Wang, Jing; Ong, Jason J.; Tang, Weiming; Zou, Maosheng; Bai, Lu; Ding, Miao; Shen, Mingwang; Zhuang, Guihua; Fairley, Christopher K

OBJECTIVES: The largely resolved first wave of the COVID-19 epidemic in China provides a unique opportunity to investigate how the early characteristics of the COVID-19 outbreak predicts its subsequent size. METHODS: We collected publicly available COVID-19 epidemiological data from 436 Chinese cities during 16(th) January-15(th) March 2020. Based on 45 cities which reported >100 confirmed cases, we examined the correlation between early-stage epidemic characteristics and subsequent epidemic size. RESULTS: We identified a transition point from a slow- to a fast-growing phase for COVID-19 at 5.5 (95% CI, 4.6-6.4) days after the first report, and 30 confirmed cases marked a critical threshold for this transition. The average time for the number of confirmed cases to increase from 30 to 100 (time from 30-to-100) was 6.6 (5.3-7.9) days, and the average case-fatality rate in the first 100 confirmed cases (CFR-100) was 0.8% (0.2-1.4%). The subsequent epidemic size per million population was significantly associated with both of these indicators. We predict a ranking of epidemic size in the cities based on these two indicators and found it highly correlated with the actual ranking of epidemic size. CONCLUSIONS: Early epidemic characteristics are important indicators for the size of the entire epidemic.