Luo Chong, Qin Xiaoli, Xie Xiaoyu, Gao Jie, Wu Yuwei, Liang Weitao, Wu Zhong
Department of Cardiovascular Surgery, West China Hospital, Sichuan University, Chengdu, Sichuan, China.
Department of Anesthesia, West China Hospital, Sichuan University, Chengdu, Sichuan, China.
Front Public Health. 2025 Apr 25;13:1507776. doi: 10.3389/fpubh.2025.1507776. eCollection 2025.
Currently, video platforms were filled with many low-quality, uncensored scientific videos, and patients who utilize the Internet to gain knowledge about specific diseases are vulnerable to being misled and possibly delaying treatment as a result. Therefore, a large sample survey on the content quality and popularity of online scientific videos was of great significance for future targeted reforms.
This study utilized normalization data analysis methods and a basic assessment scale, providing a new aspect for future research across multiple platforms with large sample sizes and for the development of video content quality assessment scales.
This cross-sectional study analyzed a sample of 331 videos retrieved from YouTube, BiliBili, TikTok, and Douyin on June 13, 2024. In the analysis of atrial fibrillation scientific videos across four social media platforms, comprehensive metrics and a basic scoring scale revealed associations between platforms, creators, and the popularity and content quality of the videos. Data analysis employed principal component analysis, normalization data processing, non-parametric tests, paired t-tests, and negative binomial regression.
Analysis of the user engagement data using a composite index revealed a significant difference in the popularity of videos from publishers with a medical background (z = -4.285, < 0.001), no aforementioned findings were found among video platforms, however, except for the Bilibili platform. As for content quality, while the difference in the total number of videos between the two groups was almost 2-fold (229:102), the difference in qualified videos was only 1.47-fold (47:32), a ratio that was even more unbalanced among the top 30% of videos with the most popularity. Notably, the overall content quality of videos from publishers without a medical background was also significantly higher (z = -2.299, = 0.02).
This analysis of atrial fibrillation information on multiple social media platforms found that people prefer videos from publishers with a medical background. However, it appeared that these publishers did not sufficiently create high-quality, suitable videos for the public, and the platforms seemed to lack a rigorous censorship system and policy support for high-quality content. Moreover, the normalized data processing method and the basic assessment scale that we attempted to use in this study provided new ideas for future large-sample surveys and content quality review.
目前，视频平台充斥着许多质量低下、未经审查的科学视频，利用互联网获取特定疾病知识的患者很容易受到误导，进而可能延误治疗。因此，针对在线科学视频的内容质量和受欢迎程度进行大规模抽样调查，对于未来有针对性的改革具有重要意义。
本研究运用归一化数据分析方法和基本评估量表，为未来跨多个平台的大样本研究以及视频内容质量评估量表的制定提供了新的视角。
这项横断面研究分析了2024年6月13日从YouTube、哔哩哔哩、TikTok和抖音检索到的331个视频样本。在对四个社交媒体平台上的心房颤动科学视频进行分析时，综合指标和基本评分量表揭示了平台、创作者与视频的受欢迎程度和内容质量之间的关联。数据分析采用主成分分析、归一化数据处理、非参数检验、配对t检验和负二项回归。
使用综合指数对用户参与度数据进行分析发现，具有医学背景的发布者发布的视频在受欢迎程度上存在显著差异（z = -4.285，<0.001），然而，除哔哩哔哩平台外，在视频平台之间未发现上述情况。至于内容质量，虽然两组视频总数的差异几乎达到2倍（229:102），但合格视频的差异仅为1.47倍（47:32），在最受欢迎的前30%的视频中，这一比例更加失衡。值得注意的是，没有医学背景的发布者发布的视频总体内容质量也显著更高（z = -2.299，= 0.02）。
对多个社交媒体平台上的心房颤动信息进行的这项分析发现，人们更喜欢具有医学背景的发布者发布的视频。然而，这些发布者似乎没有充分为公众制作高质量、合适的视频，而且平台似乎缺乏对高质量内容的严格审查制度和政策支持。此外，我们在本研究中尝试使用的归一化数据处理方法和基本评估量表为未来的大样本调查和内容质量审查提供了新的思路。
