Exploring patterns in online discussions into the lingering impact of COVID-19,  two years on

The purpose of this study was to explore underlying patterns in the users’ discussions in an online community on the darker effects of COVID-19. Understanding these patterns is critical as they can provide new information in tailoring support to individuals facing specific post-pandemic issues. Two methods were used to identify patterns in large volumes of publicly available responses (n = 23,957 posts; ~ 1,061,825 words) from an online community. Qualitatively (using Gioia methodology), 1000 random responses were manually coded by two coders and vetted by an investigator. As it was difficult to manually code such a big dataset, a quantitative approach building a topic model was employed with a language model. Qualitative analyses revealed 20 themes, including mental health (13%), impacts of direct and indirect deaths on socio-economically vulnerable groups (e.g. children and elderly, 10.4%), increasing sociopolitical divide and vaccination debate (6.8%), and work-related issues (e.g. burnout and layoffs, 6%). Topic analyses resulted in similar categories (n = 30), including physical health, loss experiences during COVID-19 and suicide; sociopolitical impact and adaptations in pandemic lifestyle; mental health and vaccination; pandemic restrictions, youth and behavioral expectations; distrust for institutions and resource scarcity; staffing issues and personal crisis; disrupted careers; and childcare challenges and economic shifts. As researchers are harnessing vast amounts of real-time human interaction data to study a variety of public health issues, our study provides insights into the specific challenges that people experienced when it became convenient to share concerns online amid an overloaded healthcare system during the pandemic.