P.067 Incidence of tissue-defined stroke and large vessel occlusion in acute stroke alerts in a non-teaching hospital system

Abstract

Background: Stroke alerts are used to triage patients with acute neurologic change for rapid imaging evaluation. CTA has been advocated to rapidly triage stroke patients for endovascular therapy. However, the yield of this approach is not well established. We evaluated the stroke alert yield in a non-teaching hospital system. Methods: A retrospective review of radiology reports for stroke alerts using PACS archive. Cases were then followed for 72 hours to determine the types of advanced imaging obtained and the findings of those studies. Results: From January to March 2014, 269 stroke alert head CTs were performed. Subsequent imaging included 128 MRIs (48%), 25 CTAs (9%) and 2 angiograms (0.7%). There were 58 (22%) tissue-defined strokes and 16 were non-lacunar (6% stroke alerts). 61% of stroke alert head CTs were negative or reported microvascular change. Other findings included large vessel occlusion (5%), intracranial stenosis (1.5%), extracranial stenosis(1.5 %), intracranial hemorrhage (9%) and masses (13%). Conclusions: Most stroke alerts were negative for tissue-defined stroke. Based on this data, universal use of CTA in the ER to triage patients with acute neurologic symptoms may not be appropriate. An updated triage system to facilitate endovascular rescue is being analyzed for changes to advanced imaging utilization and yield.