Abstract TP195: Bridging the Gap: Leveraging AI Technology&Partnerships to Meet DIDO Benchmarks for LVO Patients

Background: Timely management of stroke patients with large vessel occlusion (LVO) eligible for endovascular reperfusion therapy (ERT) is essential for optimal outcomes. In 2023, our facility’s average Door-In Door-Out (DIDO) time was 202 minutes, exceeding the Joint Commission benchmark of ≤120 minutes. Delays stemmed from lag in CTA result notification due to reliance on radiologist interpretation, inefficiencies between LVO identification and transfer decision, and a cumbersome transfer process requiring multiple calls to thrombectomy-capable centers. To address these issues, we launched a quality improvement initiative aimed at reducing DIDO times to ≤120 minutes, improving efficiency in LVO detection and decision-making through AI automation, strengthening communication with a Comprehensive Stroke Center (CSC), and ensuring only appropriate patients were transferred for ERT. Methods: Viz.ai software was integrated with our PACS system and stroke workflow to enable real-time imaging analysis and automated LVO alerts. Multidisciplinary training for ED providers, teleneurologists, and radiologists emphasized AI-driven workflow use and faster decision-making. In partnership with a CSC, we developed standardized transfer protocols and direct communication pathways to minimize delays. Continuous improvement was supported by case reviews, DIDO tracking dashboards, and workflow evaluations to identify and address bottlenecks. Results: Following Viz.ai implementation and CSC collaboration in early 2024, our average DIDO time dropped from 202 minutes to 113 minutes, meeting the benchmark. Time from CTA completion to LVO identification was reduced by 84%, from 45 minutes to 7 minutes. Time from LVO diagnosis to transfer decision improved by 70%, from 30 to 9 minutes. These gains resulted from automated alerts that bypassed delays from manual interpretation. Standardized transfer protocols reduced coordination time and eliminated multiple phone calls. Transfers were expedited, and unnecessary ones avoided, ensuring access to ERT for appropriate candidates. Conclusion: Integrating Viz.ai and partnering with a CSC significantly improved the timeliness and efficiency of LVO stroke care. These efforts reduced DIDO times, streamlined transfer processes, and created a scalable model for delivering timely stroke treatment aligned with national standards.