Tobacco researchers have used social media data to examine tobacco industry marketing practices (e.g., influencers), and to document user experience with tobacco products. This study summarized the literature that analyzed tobacco-related social media data, including domain, social media platform, tobacco product type, and themes of findings, among other variables.
PubMed, PsycINFO, Web of Science, and Communication Source were searched between 2004 and 2022. Peer-reviewed articles were included if they were written in English, included at least one tobacco-related term, and one social media-related term, and analyzed a social media post. Two coders screened all titles and abstracts. The final sample consisted of (n=255) articles. Studies were coded for domain, social media platform, tobacco product type, data source, type of data, coding and analytic method, and presence of validation procedure, among other variables.
A total of 10,504,820,581 tobacco-related social media posts were assessed across 255 studies. User experience (54.1%) and promotion (23.1%) were the most researched domains. Researchers used data from Twitter the most (42.7%). Text (43.1%) was the most common type of data analyzed. Thematic analysis (80.8%) was the most common analytic technique. Themes of findings from content analyses often pertained to the health effects of tobacco use (61.0%) and promotion (44.2%).
Researchers have analyzed billions of tobacco-related social media posts to describe user experience with, and promotions related to, tobacco products like e-cigarettes on platforms like Twitter. Future research may examine tobacco-related social media data from newer platforms like TikTok.
Real-time surveillance of tobacco-related content on social media can keep the tobacco control community abreast of tobacco industry promotional strategies, user experience with tobacco products, and perceived health effects of tobacco use. A framework may be developed to establish best practices for social media data collection and analysis, including strategies to identify posts from bot accounts and validate methodological approaches used in thematic analysis.