Introduction/Purpose Identification of large vessel occlusion (LVO) has become increasingly important with the recent publications of five favorable acute stroke thrombectomy trials. Many LVO screening scales have arisen, but none have been prospectively evaluated on the same patients to see which is most accurate.

Methods VAN, a screening tool designed based on neurovascular anatomy was published in JNIS last year. We used the prospectively collected NIHSS to derive RACE (Rapid Arterial oCclusion Evaluation Scale), FAST ED (Field Assessment Stroke Triage for Emergency Destination), and CPSSS (Cincinnati Pre-hospital Stroke Severity Scale) on the same patients and had our Nurses perform VAN exam since it is the only one not based solely on the NIHSS questions. The accuracy of RACE, FAST ED, CPSS, and VAN for identification of LVO was analyzed for positive predictive value (PPV), sensitivity, negative predictive value (NPV), specificity, and overall accuracy. Stroke mimics where not removed as this would reflect real life use.

Results 303 acute stroke codes were activated during the screening period. 36 had LVOs. RACE, FAST ED, CPSSS, and VAN each had an overall accuracy of 89.4%, 92.4%, 90.4%, and 96.7% respectively. VAN out performed all other tools in PPV, Sensitivity, NPV, Specificity, and Accuracy. VAN consistently identified LVOs in patients with NIHSS less than ten which no other tool did consistently.

RACE Rapid Arterial oCclusion Evaluation Scale, FAST ED: Field Assessment Stroke Triage for Emergency Destination, VAN: Vision, Aphasia, Neglect, CPSSS: Cincinnati Pre-hospital Stroke Severity Scale,

Conclusion In this large cohort, the VAN screening tool has been validated and identified LVO more accurately than any other screening tool. This suggests that severity scoring is less accurate than the type of stroke screening. A larger study to reproduce validation in the ED and in the field for triage is warranted.

Disclosures M. Teleb: None. A. Ver Hage: None.