Quality and Reliability Assessment of Scarlet Fever-Related Videos on Bilibili and TikTok: A Cross-Sectional Study
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 Scarlet fever remains a significant childhood infectious disease globally. Online-video platforms like Bilibili and TikTok have become vital sources of health information, yet their quality and reliability for scarlet fever content are unassessed. Objective This study evaluates the quality, reliability, and content coverage of scarlet fever-related videos on Bilibili and TikTok. Methods A cross-sectional analysis of 175 videos (95 from TikTok, 80 from Bilibili) was conducted using the keyword "猩红热" (scarlet fever). Videos were assessed using the Global Quality Scale (GQS), modified DISCERN (mDISCERN), and JAMA benchmark tools. Content coverage (epidemiology, etiology, clinical manifestations, diagnosis, treatment, prevention, transmission mode, prognosis, complication) and uploader categories (medical professionals, individual users, institutions) were analyzed. Results TikTok videos showed significantly higher engagement metrics (likes, collections, comments, shares; p < 0.001) but shorter duration (median 68s vs. 121s, p < 0.001) compared to Bilibili. Medical professionals uploaded 52.57% of videos, with TikTok hosting more professional content (68.42% vs. 33.75%). Videos by medical professionals scored higher in GQS (median = 2), mDISCERN (median = 2), and JAMA (median = 3) than individual uploaders (p < 0.01). Clinical manifestations were the most covered topic (83.43%), while prognosis was rarely addressed (12%). GQS correlated weakly with collections (r = 0.19), and JAMA scores correlated with engagement metrics (r = 0.34–0.50). Conclusion Scarlet fever-related videos on TikTok and Bilibili exhibit suboptimal quality and significant content gaps. Medical professionals produce more reliable content, but platform algorithms and user engagement do not consistently reflect quality. Enhanced content moderation and creator authentication are recommended.