Multidimensional assessment of quality, reliability, and engagement-quality discordance in ultrasound-guided nerve block videos on Bilibili (2023–2025): a cross-sectional study

There is a paucity of training resources for ultrasound-guided nerve blocks (UGNB). Although Bilibili is a widely utilised platform by trainees, the quality of its UGNB videos is unclear. In this study, the top 100 UGNB-related videos ranked by default relevance were retrieved from Bilibili on 8 January 2026, covering publications from 2023 to 2025. We included a total of 258 videos in the final analysis. No single instrument fully assesses the reliability, transparency, overall educational quality, and accuracy specific to the procedure of UGNB videos, each video was independently assessed by two reviewers using four validated, complementary tools: the modified DISCERN (mDISCERN), the Medical Quality Video Evaluation Tool (MQ-VET), the Journal of the American Medical Association (JAMA) benchmark criteria, and the Block-Specific Quality Score (BSQS). Inter-rater agreement was assessed with Cohen’s kappa and intraclass correlation coefficients. Non-parametric tests and Spearman’s rank correlation examined temporal trends and associations. Scores across all four instruments increased over time, with 2025 videos generally rated higher than those from earlier years. Videos uploaded by academic institutions scored significantly higher than those from individual creators on mDISCERN, MQ-VET, JAMA, and BSQS, with all comparisons reaching statistical significance ( P < 0.0001 for each). Engagement metrics (view count, likes, and Video Power Index) were moderately to strongly intercorrelated, with correlation coefficients ranging from 0.5 to 1.0 ( P < 0.05 for all pairwise comparisons) but showed weak-to-moderate negative correlations with quality scores ( r = − 0.5 to 0, P < 0.05), no positive correlation was detected. Videos citing references or presented in English tended to score higher. UGNB videos on Bilibili are heterogeneous but improving. Institutional content demonstrated greater reliability, whereas popularity did not reliably reflect educational quality.