Abstract WP087: Gaps in Endovascular Thrombectomy for Patients Transferred with Large Vessel Occlusion Stroke

Introduction: Patients residing in disadvantaged neighborhoods experience disparate stroke care. For patients with large vessel occlusion stroke (LVOS) presenting directly to an endovascular thrombectomy (EVT)-capable center, higher Area Deprivation Index (ADI) can result in decreased EVT. In neurologically underserved areas, patients with LVOS are transferred for EVT evaluation. HYPOTHESIS: Among patients with LVOS transferred to EVT-capable centers, higher ADI leads to lower likelihood of undergoing EVT. Methods: In stroke system registry in Texas, we identified patients with LVOS transferred to one of four EVT-capable centers (1/2018-10/2020). We assessed occurrence of EVT based on ADI quartile. The multivariate logistic regression analysis examining the odds of undergoing EVT included patient demographics (age, sex, race and ethnicity: non-Hispanic White (NHW), non-Hispanic Black (NHB), Hispanic, and other), baseline clinical characteristics (initial NIHSS and mRS, stroke risk factors: heart failure, prior stroke, hypertension, atrial fibrillation, and carotid disease), ground distance, and perfusion mismatch ratio (cutoff ≤1.2). Results: Of 518 patients transferred with LVOS (Table 1), 48% were female, and 49% were NHW, 22% NHB, and 19% Hispanic, with median age 69 years (range 19-96). Patients had a median baseline mRS of 0 (Q1-Q3 0-1) and NIHSS of 14 (Q1-Q3 6-20), residing a median of 51 (range 2-285) miles from the accepting hospital. 12 patients (2.3%) had a perfusion mismatch ratio ≤1.2. Overall, nearly half (271, 52%) underwent EVT. Patients with HTN (OR 0.58, CI 0.38-0.88), lower perfusion mismatch (OR 0.28, CI 0.06-1.18), and increased mRS (OR 0.37, CI 0.23-0.59) were less likely to undergo EVT, and patients with increased NIHSS more likely (OR 7.85, CI 4.50-13.68) (Table 2). EVT receipt did not significantly differ across ADI quartiles. Hispanic patients were less likely to undergo EVT as compared with NHW (OR 0.54, CI 0.29-0.89), regardless of ADI, age, sex, mRS, NIHSS, stroke risk factors, perfusion mismatch, and distance. CONCLUSIONS: Area deprivation and distance to EVT-capable centers did not affect EVT occurrence among patients transferred with LVOS. While sample size may limit interpretation, our findings may reflect use of standardized rapid EVT transfer and prenotification protocols within a system. Further study is warranted to understand why Hispanic patients experienced less EVT, even when adjusting for factors commonly limiting EVT.