Application of a Language Model Tool for COVID-19 Vaccine Adverse Event Monitoring Using Web and Social Media Content: Algorithm Development and Validation Study
2024

Monitoring COVID-19 Vaccine Side Effects Using Social Media

publication Evidence: moderate

Author Information

Author(s): Mackey Tim, Liu Xiaoyu, Liew Kongmeng, Scaboro Simone, Daluwatte Chathuri, Khromava Alena, Chen Yuning, Serradell Laurence, Chabanon Anne-Laure, Chan-Ou-Teung Anthony, Molony Cliona, Juhaeri Juhaeri

Primary Institution: Sanofi

Hypothesis

Can social media data be used to monitor adverse events related to COVID-19 vaccines in real time?

Conclusion

The study shows that monitoring social media can help identify adverse events related to COVID-19 vaccines more quickly than traditional reporting systems.

Supporting Evidence

  • Most adverse events were identified in 2021, with fewer in 2022.
  • The analysis showed that social media data can provide real-time insights into vaccine safety.
  • Adverse events observed were consistent with those reported by health authorities.
  • The study utilized a language model to enhance the detection of adverse events.
  • Social media monitoring can help adjust signal detection strategies for vaccine safety.
  • Data was collected from 190 countries in 61 languages.
  • The Soteria web app allows for visual analytics of adverse events.
  • Findings suggest that social media can augment traditional pharmacovigilance methods.

Takeaway

This study looks at how we can use social media to find out if people are having problems after getting COVID-19 vaccines, which can help keep everyone safe.

Methodology

A language model-based web app was developed to analyze social media and online content for adverse events related to COVID-19 vaccines.

Potential Biases

There is a risk of bias as not all patients are aware of reporting systems, and social media may not capture all adverse events accurately.

Limitations

The study may have country bias due to varying social media usage and potential over-reporting of adverse events.

Digital Object Identifier (DOI)

10.2196/53424

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication