JMIR Publications

ABSTRACT

Background:

The initiation of clinical trials for mRNA HIV vaccines in early 2022 revived public discussion about HIV vaccines, which had seen over three decades of unsuccessful research. These clinical trials followed the success of mRNA technology in COVID vaccines but unfolded in the backdrop of intense vaccine debates during the COVID pandemic. In light of the context, it is crucial to gain insights into public discourse and reactions about potential new vaccines, and social media like the X platform provide important channels.

Objective:

This study aimed to investigate the patterns of public discourse and the message-level drivers of user reactions on the X platform regarding HIV vaccines through the analysis of posts using machine learning algorithms. We examined how users employed different post types to contribute to topics and valence, and how these topics and valence influenced like and repost counts. Additionally, the study intended to identify prominent anti-HIV vaccine conspiracy theories through machine learning and manual coding.

Methods:

We collected English-language original posts about HIV vaccines on the X platform from January 1, 2022 to December 31, 2022 (N=36,424). We employed topic modeling and sentiment analysis to uncover latent topics and valence, which were subsequently used in cross-tabulation analyses to investigate their distribution across different post types. Additionally, topics and valence were integrated into linear regression models to predict user reactions, specifically like and repost counts. Finally, we manually coded 1,000 negative posts that received the most reactions to identify prominent anti-HIV vaccine conspiracy theories.

Results:

Topic modeling revealed three topics. The primary topic was HIV and COVID, followed by mRNA HIV vaccine trials, and HIV vaccine and immunity. The overall valence of the posts was marginally positive. Compared to self-composed posts that initiate new conversations, quote posts and replies, which contribute to existing conversations, featured a higher proportion focusing on HIV and COVID, and contained a greater proportion of negative posts. The topic of mRNA HIV vaccine trials, which was most evident in self-composed posts, had a positive influence on like counts. Positive valence also increased like counts. Notably, prominent anti-HIV vaccine conspiracy theories often involved linking HIV vaccines with concurrent COVID and other HIV-related events into false relationships.

Conclusions:

The results highlight COVID as a significant context for public discourse and reactions regarding HIV vaccines, in both positive and negative perspectives. The success of mRNA COVID vaccines shed a positive light on HIV vaccines. However, COVID also situated HIV vaccines in a negative context, as observed in some anti-HIV vaccine conspiracy theories that misleadingly connected HIV vaccines with COVID. These findings have implications for public health communication strategies concerning HIV vaccines.