ABSTRACT NUMBER: ESOC2026A2326 HOW A DEDICATED COMMUNICATION TOOL INTEGRATED WITH AI IMAGE INTERPRETATION ELIMINATED COMMUNICATION ERRORS IN THE NORTH OF SCOTLAND THROMBECTOMY PATHWAY

Abstract Background and aims Endovascular thrombectomy (EVT) is a time-critical intervention dependent on rapid, reliable multidisciplinary coordination. Communication failures are a recognized contributor to treatment delay and patient harm. Widely used generic messaging platforms lack stroke-specific workflows, security, and structured information exchange. We aimed to reduce communication errors and optimize decision-making within the North of Scotland EVT pathway. Methods This study evaluated the first European implementation of a secure, stroke-specific communication platform (Pulsara). It was implemented across the regional EVT network and integrated with automated CT interpretation software (Brainomix). The system supports structured, patient-centred information sharing and real-time dissemination of imaging outputs among pre-hospital, stroke, interventional, and anesthetic teams. Pathway-wide multidisciplinary training preceded implementation. Documented communication errors were compared between the pre-implementation period (2020–2023) and post-implementation period (2023–2025). Results Before implementation, EVT decision-making required a mean of 27 telephone calls per case and five clinically significant communication errors were identified, including unreported clinical deterioration during transfer and omission of antiplatelet use. Following implementation, information exchange was consolidated into a single secure patient channel with immediate access to AI-supported imaging. No communication errors were recorded during the post-implementation period, representing a significant reduction compared with baseline (P < 0.05). Conclusions Deployment of an integrated digital communication and AI imaging solution eliminated communication errors within a complex EVT pathway. This system-level intervention enhances patient safety and provides a transferable model for other time-critical emergency services, including myocardial infarction and major trauma. Conflict of interest All authors: nothing to disclose.