1118. In-House Ischemic Stroke and the Impact of Co-Morbidities on Thrombectomy Outcomes

Authors: David Lee Dornbos III, MD; Joshua Wang; Jeeho Kim; Shahid Nimjee, MD, PhD; Ciaran Powers, MD, PhD (Columbus, OH)

Introduction: The benefit of mechanical thrombectomy is well studied, with time to intervention being a significant prognostic factor. We sought to evaluate stroke outcomes among patients developing a stroke while currently hospitalized compared to those presenting from the community. Methods: We performed a retrospective analysis of all patients undergoing thrombectomy from January 2015 through December 2017. We evaluated baseline demographics, co-morbidities, NIHSS, procedural metrics and outcomes between patients presenting in-house (22 patients) or from the community (92 patients). Premorbid medical condition was assessed using the Charlson Comorbidity Index (CCI). Statistical analysis was performed using independent-samples t-test, univariate logistic/linear regression, and multivariable regression analysis where appropriate. Results: There were no significant differences in demographics, stroke metrics, NIHSS scores, or procedural data. Among patients presenting in-house, time to intervention was modestly faster (3.8±1.3 hrs v 5.3±0.3 hrs, p=0.336) and CCI score was slightly higher (5.5±0.7 v 3.9±0.2, p=0.104), although these were not significant. Patients presenting from the community had lower 90-day mortality (19.8±4.3% v 59.1±10.7%; p<0.01), decreased 90-day modified Rankin Scale (mRS) scores (3.3±0.2 v 3.7±0.5, p<0.05), and increased functional independence (mRS≤2) at 90 days (32.6±5.1% v 22.7±9.1%, p<0.05). Higher CCI scores correlated with disposition to a skilled nursing facility or hospice, rather than home or a rehab (OR 0.734, 95% CI 0.615-0.876, p<0.001), increased mortality (OR 1.359, 95% CI 1.115-1.656, p<0.01), worsened mRS scores (β 0.363, 95% CI 0.147-0.435, p<0.001), and decreased functional independence (OR 0.801, 95% CI 0.665-0.966, p<0.05). When controlling for CCI score as a potential confounder, in-house stroke presentation had no impact on mortality or functional independence. Conclusion: Patients undergoing thrombectomy for in-house ischemic stroke experience worsened long-term mortality and functional independence, despite slightly faster time to intervention. This difference appears to be attributable to worsened pre-morbid medical status, which has a substantial impact on mechanical thrombectomy outcomes.