ABSTRACT
AIMS: This study aimed to explore the experience of individuals who claimed to be COVID-19 positive via their Twitter feeds. BACKGROUND: Public social media data are valuable to understanding people's experiences of public health phenomena. To improve care to those with COVID-19, this study explored themes from Twitter feeds, generated by individuals who self-identified as COVID-19 positive. DESIGN: This study utilized a descriptive design for text analysis for social media data. METHODS: This study analysed social media text retrieved by tweets of individuals in the United States who self-reported being COVID-19 positive and posted on Twitter in English between April 2, 2020, and April 24, 2020. In extracting embedded topics from tweets, we applied topic modelling approach based on latent Dirichlet allocation and visualized the results via LDAvis, a related web-based interactive visualization tool. RESULTS: Three themes were mined from 721 eligible tweets: (i) recognizing the seriousness of the condition in COVID-19 pandemic; (ii) having symptoms of being COVID-19 positive; and (iii) sharing the journey of being COVID-19 positive. CONCLUSION: Leveraging the knowledge and context of study themes, we present experiences that may better reflect patient needs while experiencing COVID-19. The findings offer more descriptive support for public health nursing and other translational public health efforts during a global pandemic.