WeChat, a Chinese social media, may early detect the SARS-CoV-2 outbreak in 2019

Wenjun Wang; Yikai Wang; Xin Zhang; Yaping Li; Xiaoli Jia; Shuangsuo Dang

The activity of SARS in WeChat was abnormally high in Dec 1 to 3, 2019 compared to the days ahead and after. Data by now shows that the symptom onset date of the first patient identified was Dec 1, 2019. Whether there was an inner link between the word activity in WeChat and early patients is unknown. The activity of novel coronavirus in WeChat was abnormally high on Dec 11, 2019 with an Index value of 400. However, its baseline value of WeChat Index (value equals 0 or 50) was very low compared with those of other keywords (Appendix A). It is unclear whether the abnormality was a signal or a noise. As for keywords related to symptoms such as shortness of breath, dyspnea, and diarrhea, WeChat Index increased in value ahead of 9-13 days of the outbreak announcement. These symptoms are not specific to SARS-Cov-2 infection. The increases may have some association with emergence of the disease, or just a coincidence. Though lack inner link and not performed that well in detecting the SARS-CoV-2 outbreak compared with keyword Feidian, other keywords explored in the study, including those not considered as potential candidates for the outbreak sign, and even keywords not explored in the study (such as the names of drugs treating SARS-CoV-2 infection), may still have value in future outbreak detection and monitoring. Experience from Google Flu Trends showed that a combination of potential keywords other than a single keyword is more successful [9] .