Quality evaluation of short videos about insomnia on TikTok and Bilibili

This study systematically assesses the content quality and reliability of short videos related to insomnia on 2 major platforms, TikTok and Bilibili, and explores the influence of platform ecology and uploader identity on information dissemination. On November 22, 2025, searches were conducted on both platforms using the keyword “insomnia.” The top 100 relevant videos under the default ranking from each platform were included (final sample: 51 from Bilibili, 79 from TikTok). Two physicians independently assessed video quality using the Global Quality Scale, the modified DISCERN (mDISCERN) scale, and the Journal of the American Medical Association benchmark criteria. Video characteristics, interaction data, and user comment texts were analyzed. The overall quality of the included videos was moderate (median Global Quality Scale = 3.0; median mDISCERN = 3.5). Platform differences were significant: Bilibili videos were longer (median duration 975 seconds) and had higher mDISCERN and Journal of the American Medical Association scores than TikTok videos (all P < .001). TikTok videos received more likes ( P < .001). Content from designated health professionals demonstrated the highest scientific quality ( P < .001). User interaction metrics (comments, shares) showed no significant correlation with video scientific quality (prediction model R 2 ≈ 0), indicating that popularity does not reflect reliability. The quality of insomnia-related short videos varies with platform ecology and uploader identity. Professional sources are key to ensuring scientific quality; however, public interaction data (likes, comments) primarily reflect emotional resonance rather than content quality. Differentiated platform governance, enhanced cross-platform narrative skills for creators, and strengthened public education on critical digital health literacy are recommended.