Bilibili and Douyin (Chinese TikTok) videos as sources of health information about ventricular septal defect: a cross-sectional content and quality analysis of top-ranked patient-facing videos.

Résumé

BACKGROUND: The rapid growth of short-form social media and video-sharing platforms has greatly expanded public access to health information; however, the characteristics and quality of videos concerning ventricular septal defect (VSD) have not yet been systematically evaluated. Although VSD is a well-established congenital heart disease, patients and caregivers may still rely on social media to understand diagnosis, prognosis, follow-up, and treatment options. In this study, the intended audience was defined as patient-facing general users, particularly patients, parents, and caregivers who may have encountered the clinical diagnosis term "ventricular septal defect" in medical consultations, ultrasound reports, or discharge summaries and subsequently searched for related explanations online. Therefore, evaluating the quality of highly visible VSD-related videos is relevant to patient education and digital health communication. This study aimed to assess the quality and reliability of short-form videos related to VSD on the Douyin (Chinese TikTok) and Bilibili platforms.METHODS: The Chinese keyword "" was used to retrieve relevant videos from Douyin and Bilibili. This clinical term was selected because the study focused on users who search after exposure to a formal diagnosis; colloquial expressions such as "heart hole" were not included because they may retrieve heterogeneous videos on congenital heart disease in general rather than VSD-specific content. This choice is acknowledged as a limitation.To capture the content most likely to be encountered by general users, the first 100 videos returned by each platform's default search order were screened, yielding 129 videos for final analysis. The Global Quality Score was defined as the primary outcome because it provides an overall assessment of educational quality and usefulness for patients. The Video Information and Quality Index, Patient Education Materials Assessment Tool, JAMA Benchmark criteria, and modified DISCERN were used as secondary complementary measures assessing audiovisual presentation, understandability/actionability, transparency, and reliability, respectively.Video eligibility was assessed independently by two reviewers, and disagreements were resolved through discussion with a third reviewer. Because reviewer-level screening decisions were not retained, Cohen's κ for eligibility screening could not be calculated.RESULTS: Within the sampled top-ranked videos, Douyin videos demonstrated significantly higher overall quality than those on Bilibili, particularly with respect to mDISCERN, GQS, and VIQI scores. Spearman correlation analysis revealed no significant association between video quality and audience engagement on Bilibili (p > 0.05). In contrast, on Douyin, PEMAT and VIQI scores showed moderate positive correlations with likes, collections, and shares (r = 0.48-0.64, p