Engagement–quality mismatch in robot-assisted total knee arthroplasty short videos on TikTok and Bilibili
摘要
Robot-assisted total knee arthroplasty (RA-TKA) is increasingly popular in modern orthopedic surgery, yet the quality and reliability of RA-TKA content on short-video platforms are not well-defined. This cross-sectional study assessed the quality and reliability of short videos related to RA-TKA on TikTok and Bilibili. A systematic search identified the top-ranked 100 videos on each platform, of which 97 videos met the inclusion criteria (68 from TikTok and 29 from Bilibili). The Global Quality Score (GQS) and the DISCERN instrument were utilized to evaluate video quality and reliability. Engagement metrics, uploader characteristics, content categories, and video features were also analyzed. TikTok videos showed higher engagement metrics, while Bilibili videos had better informational quality and reliability, reflected in higher median GQS and DISCERN scores. Videos created by professionals and institutions generally received higher quality scores compared to those by nonprofessionals, though the differences were not always statistically significant. Content focusing on surgical treatment, technology and equipment, and balanced discussions of advantages and limitations showed higher quality and reliability than promotional content. Correlation analysis revealed that engagement metrics were strongly interrelated but did not reliably predict information quality, while longer video duration was associated with higher GQS and DISCERN scores. Videos from professional sources tended to receive higher quality scores, although differences across uploader categories were not consistently statistically significant. These findings suggest a mismatch between popularity and informational value on short-video platforms and suggest that the growing clinical uptake of RA-TKA may not yet be matched by an equally visible supply of structured, balanced public-facing information. Greater professional engagement is warranted to enhance the accurate dissemination of complex robotic orthopedic information in short-video environments.