Multidimensional quality and reliability of aortic dissection videos on Douyin and Bilibili: a cross-sectional study

Abstract Background Aortic dissection is a time-critical emergency, and short-video platforms may shape symptom recognition and care seeking. We compared the content, understandability, quality, and reliability of prominently ranked videos on two Chinese platforms. Methods In this cross-sectional study, the first 100 default-ranked results for the Chinese term for aortic dissection were screened on Douyin and Bilibili on 22–24 November 2025. Eligible videos were assessed with PEMAT-A/V (understandability only), VIQI, GQS, and modified DISCERN. Content was coded for symptoms, risk factors, diagnosis, treatment, and prognosis. Platform comparisons were descriptive and unadjusted. Scale-appropriate inter-rater agreement was estimated with 95% confidence intervals (CIs). A separate July 2026 search examined three symptom terms. Results Of 200 results screened, 170 videos were included (Douyin, 80; Bilibili, 90). Overall medians were 77% (interquartile range [IQR], 62–85) for PEMAT-A/V, 22 (IQR, 19–23) for VIQI, 3 (IQR, 3–4) for GQS, and 3 (IQR, 3–4) for modified DISCERN. Total scores did not differ clearly by platform (all P ≥ 0.072). Diagnosis appeared more often in Bilibili videos (68.9% vs 47.5%; odds ratio [OR], 2.45; 95% CI 1.31–4.58), whereas prognosis appeared more often on Douyin (30.0% vs 12.2%; OR, 0.32; 95% CI 0.15–0.72). Only 13 videos (7.6%) covered all five domains. Quality-engagement correlations were weak (Spearman rho, − 0.24 to 0.04). Agreement estimates were 0.814–0.861. In the later search, 85 of 600 results (14.2%) explicitly concerned aortic dissection. Conclusions The prominently ranked videos formed a heterogeneous information environment: overall quality scores were similar across platforms, complete clinical coverage was uncommon, and popularity was not a useful proxy for quality. These findings describe a time-specific search snapshot rather than intrinsic platform performance. Public-facing videos should prioritise early warning signs, urgent action, diagnostic pathways, and prognosis while retaining clear source attribution.