No abstract

Introduction: Large Vessel Occlusion (LVO) makes up about 30% of acute strokes. There are several published tools designed for detection of LVO for use in the prehospital setting. The ideal prehospital LVO tool should be simple and easy for prehospital staff to use and result in few LVO patients being missed as well as prevent over-testing and unnecessary transferring of patients without an LVO. Methods: From January 2015 to December 2016 1,108 acute ischemic stroke cases were identified; cases which did not present in the acute (<24hrs from onset) period and cases which did not have angiography (CTA, MRA, or catheter angiogram) performed were excluded, leaving 973 cases for analysis. The components of the initial NIHSS evaluation was utilized to calculate various prehospital LVO scales (PASS, VAN, RACE, FAST-ED, and CPSSS). Each chart was reviewed for the presence or absence of LVO defined as M1, M2, ICA or basilar occlusion. The relative performance of each of the prehospital LVO scales were compared using 2x2 analysis. Results: In the study population, 30% were LVO positive, 24% received IV tPA, and 9% received endovascular thrombectomy. Comparison of tools using previously defined cutoff points yielded the sensitivity and specificity characteristics depicted in table 1 which are compared to the NIH Stroke Scale cutoff point of 6 which has been previously defined as optimal for LVO detection. Conclusion: In this analysis, it is evident that there are distinct trade-offs, with no tool being superior when it comes to having both optimal sensitivity and specificity. Although the NIHSS is the most sensitive tool, use of this scale is not practical in the prehospital setting. Therefore, we support the use of simple tools such as PASS and VAN given the relative ease with which these tools can be learned and applied by EMS.