Assessment of the quality and reliability of short-form videos on subarachnoid hemorrhage: a cross-platform content analysis

Abstract Short-form video platforms are increasingly used for health information, but the quality and reliability of subarachnoid hemorrhage (SAH) content remain unclear. This study evaluated Chinese-language SAH-related videos across three major platforms. This cross-sectional study analyzed 228 videos from Douyin, Bilibili, and Xiaohongshu. Video characteristics, engagement metrics, uploader type, and six predefined content domains were recorded. Information quality, reliability, and transparency were assessed using the Global Quality Score (GQS), modified DISCERN (mDISCERN), and JAMA benchmark criteria. Nonparametric comparisons, effect sizes, false discovery rate (FDR) correction, correlations, inter-rater reliability, and platform-adjusted analyses were performed. Median scores were 3.0 (IQR 3.0–4.0) for GQS, 2.0 (2.0–3.0) for mDISCERN, and 1.0 (1.0–1.0) for JAMA. Video duration and engagement differed significantly across platforms after FDR correction (all adjusted P < 0.001). mDISCERN differed across platforms (adjusted P < 0.001), whereas GQS did not ( P = 0.730) and the nominal JAMA difference did not remain significant after FDR correction (adjusted P = 0.061). Clinical manifestations (79.8%) and etiology (72.4%) were most frequently covered, while prevention (12.7%) and epidemiology (11.8%) were underrepresented. Uploader-related differences in mDISCERN were attenuated after platform adjustment. Engagement metrics were strongly intercorrelated (ρ = 0.71–0.88) but weakly associated with quality measures (|ρ| ≤ 0.264). Inter-rater reliability was high (weighted κ = 0.878–0.928). Chinese-language SAH videos varied markedly in presentation and engagement, while differences in information quality were more limited. Platform effects may partly explain apparent uploader-related differences, and engagement should not be considered a proxy for information quality.