BACKGROUND
Given the permeation of social media throughout society, rumors spread faster than ever before, which significantly complicates government responses to public health emergencies such as the COVID-19 pandemic.
OBJECTIVE
We aimed to examine the characteristics and propagation of rumors during the early months of the COVID-19 pandemic in China and evaluated the effectiveness of health authorities’ release of correction announcements.
METHODS
We retrieved rumors widely circulating on social media in China during the early stages of the COVID-19 pandemic and assessed the effectiveness of official government clarifications and popular science articles refuting those rumors.
RESULTS
We show that the number of rumors related to the COVID-19 pandemic fluctuated widely in China between December 1, 2019 and April 15, 2020. Rumors mainly occurred in 3 provinces: Hubei, Zhejiang, and Guangxi. Personal social media accounts constituted the major source of media reports of the 4 most widely distributed rumors (the novel coronavirus can be prevented with “Shuanghuanglian”: 7648/10,664, 71.7%; the novel coronavirus is the SARS coronavirus: 14,696/15,902, 92.4%; medical supplies intended for assisting Hubei were detained by the local government: 3911/3943, 99.2%; asymptomatically infected persons were regarded as diagnosed COVID-19 patients with symptoms in official counts: 322/323, 99.7%). The number of rumors circulating was positively associated with the severity of the COVID-19 epidemic (ρ=0.88, 95% CI 0.81-0.93). The release of correction articles was associated with a substantial decrease in the proportion of rumor reports compared to accurate reports. The proportions of negative sentiments appearing among comments by citizens in response to media articles disseminating rumors and disseminating correct information differ insignificantly (both correct reports: χ<sub>1</sub><sup>2</sup>=0.315, <i>P</i>=.58; both rumors: χ<sub>1</sub><sup>2</sup>=0.025, <i>P</i>=.88; first rumor and last correct report: χ<sub>1</sub><sup>2</sup>=1.287, <i>P</i>=.26; first correct report and last rumor: χ<sub>1</sub><sup>2</sup>=0.033, <i>P</i>=.86).
CONCLUSIONS
Our results highlight the importance and urgency of monitoring and correcting false or misleading reports on websites and personal social media accounts. The circulation of rumors can influence public health, and government bodies should establish guidelines to monitor and mitigate the negative impact of such rumors.
On collaborative reinforcement learning to optimize the redistribution of critical medical supplies throughout the COVID-19 pandemic
