Tracking Long COVID Discourse on Twitter: A Comparative Study across Canada, the US, and English-Speaking Europe (including the UK), Using Topic Modeling and Sentiment Analysis (Preprint)

Abstract

Social media serves as a vast repository of data, offering insights into the perceptions and emotions surrounding significant societal issues. Amidst the COVID-19 pandemic, Long COVID has emerged as a chronic health condition, profoundly affecting numerous lives and livelihoods. Given the dynamic nature of Long COVID and our evolving understanding of it, effectively capturing people's sentiments and perceptions through social media becomes increasingly crucial. By harnessing the wealth of data available on social platforms, we can better track the evolving narrative surrounding Long COVID and the collective efforts to address this pressing issue.

We aimed to investigate people’s perceptions and sentiments around Long COVID in Canada, the US, and English-speaking Europe, including the UK.

We analyzed Long COVID–related tweets from 2021 using topic modeling and sentiment analysis, and interpreted the results with public health experts. We compared timelines of topics discussed across the three English-speaking regions, and also explored sentiments associated with the tweets.

Topic modeling identified five topics, and public health experts provided interpretations of the topics based on visualization, top-ranked salient terms, and representative tweets for each topic. The interpretation for each topic was not entirely distinct from one topic to the next, resulting in difficulty identifying clear patterns when analyzing the prominence of these topics over time. However, there appeared to be an overall increasing trend in the topic of duration and suffering associated with Long COVID across the regions, consistent with the increasing burden of and awareness around Long COVID as the pandemic progressed and the number of people living with Long COVID increased.

Analysis aided by natural language processing techniques and domain expert input has the potential to produce useful insights for public health monitoring and action. This study explored the use of topic modeling and sentiment analysis on Long COVID–related tweets in Canada, the US, and English-speaking Europe. This kind of information could serve as one input in public health monitoring and surveillance systems by providing a high-level understanding of the frequency and predominance of discussions and sentiments pertaining to public health issues such as Long COVID, as well as to future efforts to address such issues.