Forecasting the COVID-19 Epidemic By Integrating Symptom Search Behavior Into Predictive Models: Infoveillance Study

Abstract

Background Previous studies have suggested associations between trends of web searches and COVID-19 traditional metrics. It remains unclear whether models incorporating trends of digital searches lead to better predictions. Objective The aim of this study is to investigate the relationship between Google Trends searches of symptoms associated with COVID-19 and confirmed COVID-19 cases and deaths. We aim to develop predictive models to forecast the COVID-19 epidemic based on a combination of Google Trends searches of symptoms and conventional COVID-19 metrics. Methods An open-access web application was developed to evaluate Google Trends and traditional COVID-19 metrics via an interactive framework based on principal component analysis (PCA) and time series modeling. The application facilitates the analysis of symptom search behavior associated with COVID-19 disease in 188 countries. In this study, we selected the data of nine countries as case studies to represent all continents. PCA was used to perform data dimensionality reduction, and three different time series models (error, trend, seasonality; autoregressive integrated moving average; and feed-forward neural network autoregression) were used to predict COVID-19 metrics in the upcoming 14 days. The models were compared in terms of prediction ability using the root mean square error (RMSE) of the first principal component (PC1). The predictive abilities of models generated with both Google Trends data and conventional COVID-19 metrics were compared with those fitted with conventional COVID-19 metrics only. Results The degree of correlation and the best time lag varied as a function of the selected country and topic searched; in general, the optimal time lag was within 15 days. Overall, predictions of PC1 based on both search terms and COVID-19 traditional metrics performed better than those not including Google searches (median 1.56, IQR 0.90-2.49 versus median 1.87, IQR 1.09-2.95, respectively), but the improvement in prediction varied as a function of the selected country and time frame. The best model varied as a function of country, time range, and period of time selected. Models based on a 7-day moving average led to considerably smaller RMSE values as opposed to those calculated with raw data (median 0.90, IQR 0.50-1.53 versus median 2.27, IQR 1.62-3.74, respectively). Conclusions The inclusion of digital online searches in statistical models may improve the nowcasting and forecasting of the COVID-19 epidemic and could be used as one of the surveillance systems of COVID-19 disease. We provide a free web application operating with nearly real-time data that anyone can use to make predictions of outbreaks, improve estimates of the dynamics of ongoing epidemics, and predict future or rebound waves.

Predicting New Daily COVID-19 Cases and Deaths Using Search Engine Query Data in South Korea From 2020 to 2021: Infodemiology Study

“…Searches related to logistical needs, particularly for “thermometer” and “mask strap,” showed higher feature effects in that period. Compared to previous studies [ 12 - 14 ], terms with higher feature effects in the models were varied, from COVID-19–related terms, symptoms, and preventive measures. For longer prediction periods, NAVER search volumes were still found to be important variables, although with lower feature effects demonstrated from values of the parameter estimates.…”

“…Searches related to logistical needs, particularly for "thermometer" and "mask strap," showed higher feature effects in that period. Compared to previous studies [12][13][14], terms with higher feature effects in the models were varied, from COVID-19-related terms, symptoms, and preventive measures.…”

“…However, limited studies are available that assess predictive performances of search volume models. Studies [12][13][14] conducted in the early months of the pandemic showed that proposed models that included search data performed better than those that did not include the search volumes. However, studies in the United States demonstrated low accuracy in model prediction [15] and variability in model performance among states and time periods [16].…”

“…A study by Rabiolo et al [ 12 ] found that models that included search data performed better than those that did not include the search volumes in the first month of outbreak prediction. Similar findings were also shown in two previous analyses—one studying data from Iran [ 13 ] and another studying data from India, the United States, and the United Kingdom [ 14 ]—for periods of 1 month and 3 months, respectively, after the first case was detected.…”

Emerging application of Google Trends searches on “conjunctivitis” for tracing the course of COVID-19 pandemic

“…Recently, however, the internet searches of other symptoms were successfully used to anticipate the second and third wave of the infections. 26 Thus, Google Trends may still represent a useful tool for the digital surveillance of COVID-19.…”