Modelling Infodemics on a Global Scale: A 30 Countries Study using Epidemiological and Social Listening Data

TL;DR

The paper tackles the global infodemic problem by modeling how epidemic activity drives information production and demand signals across 30 countries. It uses a fixed-effects panel framework with sources from WHO, OxCGRT, WHO-EARS, and Google Trends to quantify the link between $C_{it}$, $D_{it}$, and $NewDocuments_{it}$, reporting an elasticity of about $0.16$ for deaths and notable foreign-burden effects. Results show that neighboring countries' epidemic burden can dominate domestic signals, and that the relationship evolves over time, being strongest in the early vaccine rollout period. The work enables cross-border infodemic monitoring, potential nowcasting, and informs integrated public health communication strategies.

Abstract

Infodemics are a threat to public health, arising from multiple interacting phenomena occurring both online and offline. The continuous feedback loops between the digital information ecosystem and offline contingencies make infodemics particularly challenging to define operationally, measure, and eventually model in quantitative terms. In this study, we present evidence of the effect of various epidemic-related variables on the dynamics of infodemics, using a robust modelling framework applied to data from 30 countries across diverse income groups. We use WHO COVID-19 surveillance data on new cases and deaths, vaccination data from the Oxford COVID-19 Government Response Tracker, infodemic data (volume of public conversations and social media content) from the WHO EARS platform, and Google Trends data to represent information demand. Our findings show that new deaths are the strongest predictor of the infodemic, measured as new document production including social media content and public conversations, and that the epidemic burden in neighbouring countries appears to have a greater impact on document production than the domestic one. Building on these results, we propose a taxonomy that highlights country-specific discrepancies between the evolution of the infodemic and the epidemic. Further, an analysis of the temporal evolution of the relationship between the two phenomena quantifies how much the discussions around vaccine rollouts may have shaped the development of the infodemic. The insights from our quantitative model contribute to advancing infodemic research, highlighting the importance of a holistic approach integrating both online and offline dimensions.

Figures (9)

• Figure 1: Normalised cumulative curves of new cases, new deaths, and new documents. Some countries exhibit coupling between the epidemiological curves and the production of new documents (e.g. Philippines, Malaysia), but in most cases the coupling is either partial (e.g. India) or absent (e.g. France).
• Figure 2: Temporal evolution of the internal elasticity associated with the number of new deaths. We fit the panel regression model specified in \ref{['eq:model_deaths']} over 6-month sliding windows, and display how the internal elasticity $\beta_1$ and its robust standard error evolve. Left, the evolution of $\beta_1$ using the number of new documents as the dependent variable, with the inset reporting the total number of new deaths across all modelled countries. Right, the evolution of $\beta_1$ using Google Trends data as the dependent variable, as a proxy for information demand rather than production. Both curves show similar behaviour and comparable values, peaking in the first half of 2021 and stabilizing around values close to 0 from the end of 2021 onward.
• Figure 3: Country-specific internal and external elasticities associated with the number of new deaths. For each country, we fit a regression model as specified in \ref{['eq:model_deaths']} and obtain an estimate of the internal elasticity $\beta_1$ and of the external elasticity $\beta_2$. (A) Comparison of $\beta_1$ and $\beta_2$ for all countries, with the vertical and horizontal dashed lines reporting the average $\beta_1$ and $\beta_2$, respectively, and the triangle indicating the values resulting from the panel regression. (B) Difference between internal and external elasticity $\beta_1 - \beta_2$ for all countries. Nicaragua was omitted from this figure because its estimated $\beta_1$ was much higher than that of all other countries.
• Figure 4: Cross-correlation between number of new cases and number of new documents. On the $x$-axis, how many days the number of documents has been shifted by, with positive (negative) values indicating that the documents time series has been moved backwards (forwards); on the $y$-axis, the sample correlation between the two time series at the given lag.
• Figure 5: Cross-correlation between the number of new cases and the number of new documents. Four paradigmatic cases related to Senegal, India, UK and Peru are displayed.