Document resource
Background Remote live-streamed training in endovascular thrombectomy (EVT) is a novel educational strategy. This study evaluated the dose–response relationship between training duration and clinical outcomes, and explored mediation pathways.Methods In a prospective cohort study, 1046 participants received remote EVT training, with 1010 completing follow-up. Locally Estimated Scatterplot Smoothing (LOESS) and segmented regression identified training thresholds. Participants were categorized into low (<25 hours), medium (25–70 hours), and high (>70 hours) training groups. Inverse probability of treatment weighting (IPTW) was adjusted for baseline differences. Mediation analysis examined the mechanisms of effect.Results A dose–response relationship was evident: outcome improvements accelerated after 25 hours and plateaued beyond 70 hours. After IPTW, patients treated by both medium and high training groups achieved significantly better 90-day outcomes compared with the low training group (medium vs low, adjusted β=1.54, 95% CI 0.94 to 2.15; high vs low, β=3.16, 95% CI 2.28 to 4.04, both P<0.001). Mediation analysis revealed that increased professional competency (25.0%), reduced recanalization time (32.8%), and higher recanalization rates (46.1%) partially explained the improved outcomes (all P<0.001). Sensitivity analyses supported these findings.Conclusion Remote live-streamed EVT training is significantly associated with improved 90-day functional outcomes. The greatest benefits were observed after 25 hours of training and diminishing returns beyond 70 hours. These gains are partly mediated by enhanced professional competency and improved recanalization outcomes, highlighting the importance of structured, high-quality remote training programs for optimizing neurointerventional care.