Effects of Antivaccine Tweets on COVID-19 Vaccinations, Cases, and Deaths

2024-06-13

arXiv.org

AI Summary

This study quantifies the causal impact of anti-vaccine social media discourse on real-world vaccination behavior and pandemic outcomes. Method: We develop a multi-state ordinary differential equation (ODE) model integrating vaccine hesitancy, exposure to anti-vaccine content, and infectious disease transmission. Using county-level U.S. vaccination rates, COVID-19 case and mortality data, and geolocated Twitter anti-vaccine tweet distributions, we perform Bayesian parameter estimation and spatial causal attribution to establish an identifiable causal inference framework linking online discourse to offline health outcomes. Contribution/Results: Between February–August 2021, a one-standard-deviation increase in county-level anti-vaccine tweet exposure significantly reduced vaccination uptake. Nationally, approximately 750,000 individuals forewent vaccination due to such exposure, leading to at least 29,000 additional infections and 430 excess deaths. This work provides the first causal evidence and computational paradigm for digital public health interventions targeting misinformation-driven vaccine refusal.

Technology Category

• Reasoning under Uncertainty: Causality
• Application Domains: Misinformation & Fake News
• Machine Learning: Causal Learning

Application Category

• Responsible Web: Mitigating misinformation and disinformation
• Social Networks and Social Media: Media and governance, opinion dynamics, filter bubbles, polarization
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

Vaccines were critical in reducing hospitalizations and mortality during the COVID-19 pandemic. Despite their wide availability in the United States, 62% of Americans chose not to be vaccinated during 2021. While online misinformation about COVID-19 is correlated to vaccine hesitancy, little prior work has explored a causal link between real-world exposure to antivaccine content and vaccine uptake. Here we present a compartmental epidemic model that includes vaccination, vaccine hesitancy, and exposure to antivaccine content. We fit the model to observational data to determine that a geographical pattern of exposure to online antivaccine content across US counties is responsible for a pattern of reduced vaccine uptake in the same counties. We find that exposure to antivaccine content on Twitter caused about 750,000 people to refuse vaccination between February and August 2021 in the US, resulting in at least 29,000 additional cases and 430 additional deaths. This work provides a methodology for linking online speech to offline epidemic outcomes. Our findings should inform social media moderation policy as well as public health interventions.

Problem

• Investigating antivaccine tweets' impact on COVID-19 vaccination rates
• Linking online misinformation to real-world health outcomes
• Quantifying causal relationship between social media content and vaccine hesitancy

Innovation

• Compartmental epidemic model with vaccine hesitancy
• Linking Twitter antivaccine exposure to vaccination rates
• Methodology connecting online speech to offline outcomes