All information used in this study was posted publicly, and the study did not involve any interaction with users. User identifiable information was removed.

The hashtag #nicsick was originally selected for review based on news and media reports on the popularity of the term for users’ self-reporting experiences and topics associated with nicotine sickness [21]. The selection of #nicsick as the hashtag for data collection was confirmed by conducting preliminary manual searches with the TikTok in-platform search function for keywords, terms, and hashtags that were used in videos discussing nicotine sickness content. In addition, we conducted preliminary Google Trends search terms analysis to assess if there were any queries or topics related to the nic sick terminology set for the study period, which did not reveal any specific terms for further keyword or hashtag data filtering. TikTok videos with the hashtag #nicsick were collected retrospectively in May 2021 by first using structured searches on the platform without any user log-in credentials, with no personal search history and with cookies disabled. The URLs and metadata (eg, date, time, username, favorites) associated with retrieved TikTok videos using the hashtag #nicsick were collected and saved using an algorithm (Python, version 3.7.0) and an automated web browser and automation tool (Selenium, version 3.141.0; SeleniumHQ). In addition, the videos were also downloaded and saved for content analysis.

Content analysis of TikTok videos collected was conducted by VP, TM, and MN. The videos were manually annotated (using binary classification—whether the video depicted nicotine poisoning or nicotine sickness or not) and then inductively coded for specific themes that emerged, such as showing active vaping, specific adverse events, or nicotine sickness experiences (Table 1). Furthermore, based on prior research studies [7,32,33] that have conducted content analysis on TikTok videos, we also coded the following metrics and associated metadata: (1) user engagement (views, likes, comments, shares, followers, verification); (2) video characteristics (duration, caption, text on screen, subtitles, music); and (3) video type (original, duet, stitch). An original video is a microvideo uploaded by a TikTok user and is the primary source of user-generated posts on the platform. A stitch integrates the user’s own video into another user's original video by clipping and integrating scenes from the original video which allows the sharing of existing videos in combination with new user-generated content. A duet video builds on another user’s original video, with a new video recorded alongside the original as it plays and is mainly used for user reaction videos. VP, TM, and MN coded posts independently and achieved high intercoder reliability (κ=0.96). For inconsistent results, senior author TKM was consulted.

After open inductive coding based on our coding schema, videos with nicotine poisoning content were grouped into main thematic categories: (1) firsthand and secondhand reporting of suspected nicotine sickness symptoms (eg, nausea, headache, cough, stomachache, and regurgitation); (2) videos displaying actual poisoning-related adverse events; and (3) videos displaying active vaping while discussing nicotine sickness–related content. The 2-tailed independent sample t test and Wilcoxon rank-sum test were used to compare the engagement metrics of videos with and without active nicotine sickness–related content. All statistical analyses were conducted using RStudio (version 3.6.1). A P value<.05 was considered statistically significant.

We found that TikTok users openly discussed their experiences, both firsthand and secondhand, with nicotine poisoning and related adverse events via the #nicsick hashtag. This includes users promoting extreme vaping behavior, discussion of users’ direct experiences with adverse events, and explicit video documentation of adverse events that appear to elicit user attention and reaction. We also found that users were more likely to create original videos as opposed to stitches and duets, which allow users to imbed and combine existing videos into their own TikTok post, which can often result in higher user engagement [34]. User engagement metrics of the videos in our study indicated moderate and lower compared to general metrics for popular TikTok videos, although some popular creators have had higher levels of engagement and mean follower counts. Additionally, the duration of videos depicting adverse events or nicotine poisoning symptoms was marginally longer than those that did not, indicating that videos depicting adverse events may need to be longer in order to provide sufficient details on the reported nicotine sickness experiences versus videos that simply reported or portrayed general vaping behavior.

These emerging forms of microvideo-based social interactions have the potential to change ways in which susceptible youth and young adults receive vaping-related behavior cues. For example, phrases heard in TikTok videos in this study included “nic-sick check” and “where’s my JUUL” and were associated with intentional overconsumption of nicotine products that also appeared to increase user engagement. TikTok’s For You Page also personalizes content based on user viewing history and can lead to a continuous cycle of exposure to trending vaping videos, including those in this study that portrayed overconsumption. Characteristics of TikTok videos, including use of hashtags, trending sounds, and video duration, may also represent crucial opportunities for the introduction of targeted health education and promotion to warn young users about the documented harms of vaping, particularly if they are already experiencing and self-reporting adverse events [35]

Yet, it is important to note that the impact of these videos on actual knowledge, attitudes, and behaviors of social media users associated with the appeal, uptake, and use of electronic nicotine delivery systems remains unclear. Vaping themes detected in this study seem to demonstrate that users may be promoting harmful vaping practices, which in turn has the potential to encourage overconsumption or dangerous nicotine use behavior. For example, popular internet challenges, including vape challenges, are rapidly circulated through TikTok—some may greatly harm the health of participants or encourage others to engage in similar behavior [25]. However, the motivations of these users to post #nicsick videos and the reason why they may or may not become viral require further study, particularly in the context of how the risk of electronic nicotine delivery system–related adverse events may be increased. We also found that there was no clear indication that users experiencing nicotine sickness were committed to quitting despite their sickness or poisoning, although a few user videos included individuals expressing a desire to quit. Instead, users using #nicsick appeared to be actively participating in trending vape tricks or challenges, which then led to overconsumption followed by one or more adverse events, which they then posted.

Combating the youth vaping epidemic requires scrutiny of how new interactive media, such as TikTok, serves as a source of vaping-related information and a source of peer influence among young user communities. Traditional poisoning and adverse event surveillance systems, such as the National Poison Data System and the US National Electronic Injury Surveillance System, may fail to capture these less severe cases captured on social media platforms, thereby underrepresenting the morbidity burden of nicotine poisoning [19]. Hence, social media platforms such as TikTok, may represent an important data source that can be used to help identify the burden of user-reported nicotine exposure and poisoning.

Our study narrowly focused on identifying and characterizing how nicotine poisoning–related content is created and shared through TikTok’s microvideo posts. Future studies should focus on examining the complex interplay between user exposure and the appeal of TikTok pro–tobacco and vaping content, the dynamics of how this content is shared and propagated on these networks, how it impacts real-world use patterns and vaping behavior, and finally, how these factors may impact safety and health issues, such as adverse events. Though TikTok’s policies state that content which offers the purchase, sale, trade, or solicitation of tobacco products (including vaping products) is prohibited [36]. Our study results indicate that content related to excessive or potentially dangerous smoking or vaping behavior does not fit this criterion, although the utility of removing such content is also unclear.

We used a single platform and a fixed data collection period. Hence, the results may not be representative of all nicotine sickness content on social media. While TikTok is a popular social media platform, the demographics of TikTok users may not reflect those of the general population of electronic nicotine delivery system users. Furthermore, study data were limited to results for videos queried and returned by TikTok’s internal search function, which may further limit generalizability. We used a single hashtag, #nicsick, based on a combination of open-source news reports, manual searches, and a preliminary Google Trends analysis to justify inclusion and due to the narrow focus of the study on nicotine sickness–related content. Hence, this study might have missed capturing additional nicotine sickness content with other hashtags that may have focused on specific electronic nicotine delivery system products, vaping behavior and challenges, and other nicotine and vaping–related topics with content related to nicotine sickness or adverse events. Furthermore, we did not cross-validate user-generated content with other data sources of nicotine sickness reports or cases, such as data from poison control centers, and we did not explore user interaction or social network characteristics of web-based user networks where videos with nicotine poisoning content was shared, viewed, or interacted with, via favorites or comments, in-depth. Future studies should conduct more comprehensive and large-scale data analysis of nicotine-related adverse events on TikTok compared with those on other social media platforms. Future research can explore further analysis of characteristics of videos including hashtags, trending sounds, and video duration, which may help curate targeted health education and promotion against vaping use.

While we observed discussions and experiences with nicotine poisoning and adverse events on TikTok, further research is needed to assess how this unique social media risk environment impacts user perceptions about the known harms of vaping and which health promotion strategies can be tailored for youth experiencing vaping-related adverse events. Although other electronic nicotine delivery system–promoting content on TikTok may influence tobacco use behavior, #nicsick-specific content represents user-generated experiences directly associated with adverse events or poisoning due to nicotine overconsumption, which could be an alternative form of surveillance to better characterize this likely underreported public health issue. Supplementing traditional surveillance methods with infodemiology and infoveillance approaches can help in capturing the nicotine sickness cases that are not reported to poison control centers and assess the growing health burden due to nicotine use more appropriately.