Abstract TMP66: Redefining Stroke Transport Patterns Using the SNOW Scale for Large Vessel Obstruction

Autor: Paul Vilar, Richard A. Rovin, Reji Babygirija, Sarah Erpenbeck, Maharaj Singh, Kessarin Panichpisal
Rok vydání: 2019
Předmět:
Zdroj: Stroke. 50
ISSN: 1524-4628
0039-2499
DOI: 10.1161/str.50.suppl_1.tmp66
Popis: Background: The SNOW (Stroke Network of Wisconsin) scale was developed to identify patients with acute ischemic stroke (AIS) due to large vessel occlusion (LVO). We previously validated the SNOW scale with two retrospective and one prospective datasets of AIS patients evaluated in the Emergency Department of Aurora Health Care System hospitals. In all three studies, the SNOW scale showed a high sensitivity and accuracy and compared favorably to other stroke scales. Based on this, Milwaukee County EMS (MCEMS) implemented a new protocol incorporating the SNOW scale to prehospital triage of AIS patients: a patient with presumed AIS with a positive SNOW score, indicating LVO, paramedics are instructed to bypass the closest stroke hospital in favor of a thrombectomy capable hospital if no more than 15 minutes is added to the transport time. This is the first report of interim analysis of the data. Methods: To be SNOW positive, a patient must exhibit at least one of the following findings: gaze deviation, expressive aphasia, or neglect. We prospectively reviewed a cohort of all suspected stroke patients MCEMS transported to the three thrombectomy capable hospitals in Milwaukee, WI, between March 1 2018 and December 31 2018. LVO was confirmed by vascular imaging and included occlusions of the intracranial internal carotid artery, middle cerebral artery (M1 and M2), anterior cerebral artery (A1 and A2), or the basilar arteries. Results: In the first four month analysis, 108 suspected stroke transports arrived to Aurora St. Luke’s Medical Center. Of these 108, 21 (19.44%) were confirmed LVO stroke patients. The MCEMS SNOW scoring correctly identified 16 cases of LVO with a sensitivity of 76.19%, specificity of 39.98%, and accuracy (area under the receiving operating characteristics curve) of 0.5764. Aurora St. Luke’s Stroke Responders SNOW scoring correctly identified 20 LVO cases with a sensitivity of 95.24%, specificity of 58.62%, and accuracy of 0.7693. Between the EMS and Hospital SNOW scoring, all confirmed LVO patients were identified. Conclusion: Upon analysis of interim data, the SNOW scale shows a high sensitivity and accuracy to predict LVO. Differences in EMS and Hospital scoring may be due to experience with administering the scale and the change in exam over time.
Databáze: OpenAIRE
Externí odkaz: