YouTube’s health content rewards attention-grabbing videos over reliable medical information

By Hugo Francisco de SouzaReviewed by Susha Cheriyedath, M.Sc.Jul 27 2026

A large study of medical YouTube videos reveals a credibility paradox: content designed to grab attention often travels farther than clearer, more reliable health information.

Study: The credibility paradox: clickbait, engagement, and information quality in YouTube’s medical ecosystem. Image Credit: simona pilolla 2 / Shutterstock

In a recent study published in the journal npj Digital Medicine, researchers investigated previously understudied relationships between clickbait packaging, viewer engagement, and medical information quality across YouTube's digital health ecosystem.

The study's dataset comprised 3,958 Korean-language medical YouTube videos, which were systematically analyzed across five major health domains. The results revealed that while clickbait strategies were associated with higher view counts and user interaction, they were also linked to lower clinical information quality.

These findings indicate that engagement-seeking platform incentives may contribute to a "credibility paradox" in which high-quality content from medical professionals and institutions remains isolated in low-visibility spaces, while independent YouTuber content with higher clickbait scores and lower quality scores attracts substantially more viewers.

Background

In today's world of almost ubiquitous internet and smartphone adoption, digital platforms are increasingly recognized as primary destinations for patient health information seeking. The Google-owned YouTube platform alone is estimated to have attracted roughly 16 billion monthly health-related video views worldwide by 2025.

While supporters of this transition welcome digital adoption's substantially reduced user costs and improved patient access to health information, experts caution that as patients increasingly rely on online media to understand disease pathophysiology and treatment protocols, the quality and accessibility of public medical education directly influence treatment adherence and health outcomes.

Furthermore, critics argue that YouTube's attention-based economy incentivizes its creators to deploy "clickbait" to compel additional user engagement. Unfortunately, whether engagement incentives are associated with lower objective clinical rigor across diverse health topics remained hitherto unquantified.

About the study

The present study aimed to address this knowledge gap and inform future discussion and policy on digital patient health information seeking by analyzing a dataset of 3,958 Korean-language YouTube videos. The study classified these videos into one of five pre-specified high-burden health topics: cancer, diabetes, weight management, depression, and dementia.

The study's evaluations first comprised custom search queries derived from 20 standardized Term Frequency-Inverse Document Frequency (TF-IDF)- filtered search terms. The study dataset, comprising YouTube videos, was evaluated by two trained independent coders using a multidimensional approach, which included three validated clinical instruments:

1. JAMA benchmark criteria, which were used to assess the content's transparency, authorship, and attribution;
2. The DISCERN instrument, which was used to evaluate treatment choice reliability, and
3. The Global Quality Scale (GQS), which served as an overall rating for the content's flow, clarity, and usefulness.

Individual instrument z-scores were averaged into a composite Medical Quality Score. Concurrently, videos were evaluated using the Clickbait Severity Score, which counted the number of clickbait tactics present in each video.

Finally, statistical analyses included Welch's t-tests, Pearson correlation matrices, and an ordinary least squares (OLS) multiple linear regression model, the latter of which was used to predict log-transformed daily views while controlling for video duration, age, and video source category.

Study findings

The study's outcomes indicate a clear hierarchy in information quality and engagement based on video source. Public and government institutions produced the most reliable and useful content, with the lowest levels of clickbait, yet averaged only 78,324.8 views.

In contrast, independent YouTubers, who made up 38.0% of the study sample, produced the lowest-rated content and used the most clickbait, yet garnered nearly 9 times as many views, averaging 674,851.9.

Clickbait prevalence analyses corroborated these findings, further identifying exaggeration (26.5%) and formatting tricks (25.0%) as the most frequently utilized clickbait tactics.

Clickbait severity showed a statistically significant positive correlation with log-transformed daily views, likes, and comments, indicating an association between clickbait severity and audience engagement.

The multiple regression model identified specific drivers of engagement: curiosity-inducing tactics, such as exaggeration and teasing, were identified as strong positive predictors of daily views. Conversely, videos rated as clearer, more useful, and higher quality tended to receive fewer daily views.

Conclusions

This study suggests that YouTube's engagement-centered visibility dynamics may favor clickbait-associated content over higher-quality medical information, forcing medical experts to choose between maintaining scientific integrity and achieving platform visibility.

While limited by its cross-sectional design and focus on Korean-language media, the study's large sample size provides a robust quantitative blueprint of a platform-level problem in health information quality. However, the authors noted that the study could not establish causality, relied on aggregate engagement metrics, lacked individual-level viewer data, and could not directly quantify YouTube's algorithmic black box.

The authors conclude that mitigating public health risks requires urgent platform governance reform, which prioritizes validated clinical quality signals over raw user engagement.