TikTok: Doctors’ SCD Videos More Accurate, But Non-Experts Get More Views

TikTok’s Algorithm Prioritizes Untrained Voices on Sudden Cardiac Death, Study Finds

A new study published Wednesday in the journal Scientific Reports reveals a significant disparity on the social media platform TikTok: videos about sudden cardiac death (SCD) created by healthcare professionals (HCPs) are demonstrably more accurate and reliable than those made by non-experts, yet the latter consistently receive far greater engagement from users.

Researchers assessed 83 TikTok videos related to SCD, analyzing their quality, sentiment, and levels of user interaction. The study, utilizing data gathered in December 2024, found that videos created by individuals without medical training garnered significantly more reshares, “favorites,” and overall engagement, despite scoring lower on assessments designed to measure the accuracy and credibility of health information.

The analysis employed three established evaluation tools: the DISCERN scale, a benchmark developed by the Journal of the American Medical Association (JAMA), and the CRAAP test – an acronym for Currency, Relevance, Authority, Accuracy, and Purpose. HCP-created videos consistently outperformed those from non-HCPs across all three metrics, demonstrating a stronger presentation of clear aims, relevant information, and verifiable sources.

“HCP videos more strongly presented clear aims, achieved their aims, showed relevance, identified sources, and clearly indicated information than non-HCP videos,” the study detailed. They also provided more comprehensive descriptions of treatment benefits, risks, and potential impacts on quality of life.

Despite the superior quality of content from medical professionals, non-HCP videos achieved higher engagement rates. The study found that non-HCP videos had a greater median number of reshares (p=0.0050) and “favorites” (p=0.0495), as well as a higher overall engagement rate (p=0.0014). This suggests that TikTok’s algorithm favors content based on popularity rather than factual accuracy when it comes to potentially life-saving health information.

Sentiment analysis revealed no significant difference in the overall emotional tone of videos created by HCPs and non-HCPs, although non-HCP videos were slightly more likely to express a positive sentiment (59.1% versus 40.9%, p=0.332). Researchers noted this difference was not statistically significant.

The study’s findings underscore growing concerns about the spread of misinformation on social media platforms, particularly regarding health-related topics. TikTok has faced increasing scrutiny for its content moderation policies and the potential for harmful content to reach a large audience. The platform’s algorithm, which prioritizes content based on user engagement, has been criticized for amplifying sensational or misleading information.

Researchers found a positive correlation between video duration and quality assessment scores, as well as between the number of “favorites” and JAMA benchmark and CRAAP test scores. However, engagement metrics like comments and likes did not show a significant association with quality. Notably, being a healthcare professional was strongly correlated with higher scores on all three quality assessments.

The study’s authors suggest that collaborations between HCPs and popular TikTok creators, or platform-level interventions designed to boost the visibility of evidence-based content, may be necessary to address the gap between information quality and audience reach. The research team indicated that the datasets used in the study will be made available upon request.