TikTok favors non-experts even when doctors post better sudden cardiac death content

By Tarun Sai LomteReviewed by Susha Cheriyedath, M.Sc.Mar 24 2026

A new study found that healthcare professionals post more credible TikTok videos about sudden cardiac death, yet non-expert creators still win more attention, exposing a gap between information quality and audience reach.

Study: Quality of TikTok videos on sudden cardiac death varies by video characteristics and health information accuracy. Image Credit: Korawat photo shoot / Shutterstock

In a recent study published in the journal Scientific Reports, researchers evaluated the quality, sentiment, and engagement of TikTok videos about sudden cardiac death (SCD).

TikTok as a Source of Health Information Risks

TikTok is one of the most popular platforms for health information among younger people. While TikTok advocates for freedom of creation and free speech, it has been criticized for inconsistently regulating false information, explicit content, and hate speech. Because content is prioritized based on engagement metrics, misleading and harmful content can go viral.

TikTok contains a large number of health-related videos predominantly created by non-medical individuals. These videos often oversimplify complex conditions and fail to adhere to evidence-based information. While the role of social media in disseminating health (mis)information has received substantial attention, studies on SCD content on TikTok are lacking.

Study Design and Video Selection Methods

In the present study, researchers analyzed the quality, sentiment, and engagement of SCD-related videos on TikTok. The Exolyt analytics platform's top video search function was used to identify SCD-related content on TikTok using two hashtags. For each hashtag, the top 50 videos were sequentially selected from ranked results. All search and data extraction tasks were performed on December 22, 2024. Duplicates and non-English videos were removed. An additional 10 videos were excluded because of insufficient data.

Content Quality Assessment and Analytical Tools

Engagement metrics, including comments, likes, and shares, were systematically reviewed. Content creators were categorized as healthcare professionals (HCPs) and non-HCPs based on publicly available self-declared profile information. Health information in videos was examined using three quality assessment tools. The DISCERN scale, a validated instrument for assessing the quality of written consumer health information, was adapted to evaluate the quality of health information and treatment presented in videos.

Further, a Journal of the American Medical Association (JAMA) benchmark that assesses health information on web pages based on attribution, authorship, currency, and disclosure was adapted to analyze videos. The Currency, Relevance, Authority, Accuracy, and Purpose (CRAAP) test, designed to assess the credibility of online sources, was adapted to evaluate videos.

A sentiment analysis was performed by reviewers to classify the emotional tone of the video content as neutral, positive, or negative. Spearman’s rank correlation coefficient was used to examine the associations between video characteristics and quality assessment scores. It was also used to analyze the relationship between sentiment scores and HCP status, quality assessment scores, number of views, follower count, video duration, and engagement rate and metrics. The paper described these analyses as exploratory.

The initial sample comprised 100 videos; after exclusions, 83 videos were selected for analysis. These included 52 non-HCP and 31 HCP videos. HCP videos were significantly longer, whereas non-HCP videos had higher reshares, favorited counts, and engagement rates. Moreover, HCP videos scored significantly higher on DISCERN, JAMA benchmark, and CRAAP assessments. More non-HCP videos (59%) showed positive sentiment than HCP videos (41%), but this difference was not statistically significant.

Detailed Quality Score Comparisons Across Videos

Among DISCERN domains, HCP videos more strongly presented clear aims, achieved their aims, showed relevance, identified sources, and clearly indicated information than non-HCP videos. In addition, HCP videos outperformed non-HCP videos in providing additional resources, balanced perspectives, and descriptions of therapeutic mechanisms, benefits, treatment options, risks, the impact on quality of life, and the consequences of no treatment.

HCP videos also showed higher scores across JAMA benchmark domains, with stronger authorship, source attribution, avoidance of promotional motives, and currency of content. Likewise, these videos scored higher in individual domains of the CRAAP test, including verifiable credentials, creator credibility, background presentation, information accuracy, avoidance of inaccuracies, objectivity, and educational purpose.

Video duration and the number of reshares were positively associated with all three quality assessment scores. Favorited videos showed a positive correlation with JAMA benchmark and CRAAP test scores. Other metrics, like engagement rate, comments, and likes, showed no significant associations. Notably, HCP status was strongly correlated with all three scores.

Sentiment was significantly negatively correlated with video duration. No significant correlations were observed between sentiment and reshares, number of views, engagement rate, and assessment scores. Moreover, sentiment did not differ by content creator type. Sentiment showed a weak positive but insignificant association with times favorited and follower count.

Implications for Health Information on Social Media

In sum, the study found that TikTok videos on SCD generated by HCPs were superior in content quality, credibility, and quality-assessment performance, but had lower engagement rates than those from non-HCPs. The higher CRAAP, JAMA benchmark, and DISCERN scores of HCP videos indicate they provide higher-quality, more credible information.

These results emphasize the need for strategies to improve the impact and accessibility of professional health information. However, the study was cross-sectional, assessed only highly ranked English-language videos under two hashtags, and relied on adapted scoring tools and reviewer judgments, so its findings should be interpreted cautiously.