Internet search data could Be used as novel indicator for assessing COVID-19 epidemic

• Main Authors: Kang Li, Yanling Liang, Jianjun Li, Meiliang Liu, Yi Feng, Yiming Shao
• Format: Article
• Language: English
• Published: KeAi Communications Co., Ltd. 2020-01-01
• Series: Infectious Disease Modelling
• Subjects: COVID-19; Internet data; ARIMAX; Predict
• Online Access: http://www.sciencedirect.com/science/article/pii/S2468042720300555

The pandemic of the coronavirus disease (COVID-19) poses a huge challenge all countries, since no one is well prepared for it. To be better prepared for future pandemics, we evaluated association between the internet search data with reported COVID-19 cases to verify whether it could become an early indicator for emerging epidemic. After the keyword filtering and Index composition, we found that there were close correlations between Composite Index and suspected cases for COVID-19 (r = 0.921, P < 0.05). The Search Index was applied for the Autoregressive Integrated Moving Average with Exogenous Variables (ARIMAX) model to quantify the relationship. Compared with the model based on surveillance data only, the ARIMAX model had smaller Akaike Information Criterion (AIC = 403.51) and the most accurate predictive values. Overall, the Internet search data could serve as a convenient indicator for predicting the epidemic and to monitor its trends.