Quality of Chronic Disease–Related Health Videos Across Social Media Platforms: A Systematic Review and Meta-analysis
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
Social media videos have become one of the major sources of health information for individuals living with chronic diseases. Although numerous cross-sectional studies have evaluated the quality of health-related videos across different platforms, the overall quality of chronic disease–related videos and the determinants underlying quality variation remain unclear.
Objective
To systematically evaluate the quality of chronic disease–related health videos across major global and Chinese social media platforms and to identify potential determinants of video quality using multivariable meta-regression.
Methods
This systematic review and meta-analysis searched PubMed, Embase, and Web of Science from database inception to April 30, 2026, for cross-sectional studies evaluating Chinese- and English-language health videos. Scores from the DISCERN instrument, the Global Quality Scale (GQS), and the Journal of the American Medical Association (JAMA) benchmark criteria were standardized to a 0–100 scale and quantitatively synthesized using random-effects models. Prespecified subgroup analyses and multivariable meta-regression were conducted to explore potential sources of heterogeneity, including platform region, platform type, disease category, video duration, professional background of content creators, and audience engagement.
Results
A total of 88 studies involving 18,688 videos were included. Overall methodological quality was suboptimal, with a pooled standardized DISCERN score of 50.61 (95% CI, 48.04–53.18), accompanied by substantial between-study heterogeneity (I² = 99.3%). Videos hosted on international platforms achieved significantly higher quality scores than those on Chinese platforms (54.68 vs. 48.21; P = 0.008). Multivariable meta-regression demonstrated that conventional predictors—including video duration, the proportion of physician creators, and audience engagement—were not independently associated with video quality ( P > 0.05). Importantly, the final model explained only 9.06% of the between-study heterogeneity (R² = 9.06%), indicating that conventional content- and creator-level characteristics account for only a small proportion of the observed variability in video quality.
Conclusions
Traditional predictors, including creator professionalism, video duration, disease category, and audience engagement, have limited ability to explain variation in the quality of online health videos. Although platform region emerged as the only significant moderator, the multivariable model explained only a small fraction of the observed heterogeneity, suggesting that the principal determinants of health information quality remain largely unexplained.
