Abstract WP58: Correlating Stroke Volume with National Institute of Health Stroke Scale as a Means of Determining Patient Prognosis

Abstract

Introduction: The NIH Stroke Scale (SS) is a widely used tool for directing treatment and predicting outcomes in Acute Ischemic Stroke (AIS). Severe strokes with high admission SS often correlate with long term disability, and as such, SS serves as a strong predictor of outcome. Final infarct volume (FIV) is also a pivotal predictor of stroke outcome. We aimed to evaluate the relationship between SS, FIV and outcome, and hypothesize that a combined approach evaluating both FIV and SS may more accurately correlate with patient outcomes. Methods: A single center, retrospective cohort study, examined AIS patients with large vessel occlusion (LVO) affecting the anterior circulation, between July 2004 and April 2013. Patients were stratified by treatment to 1) intra-arterial therapy, 2) IV tPA, 3) both or 4) neither. Primary outcomes measured were mRS at discharge and 90 days (good outcome mRS 0-2, poor 4-6). FIV was manually calculated from DWI obtained within the first 7 days of presentation. SS and FIV were compared against good and poor mRS outcomes using Wilcoxon rank sum test. Logistic regression analysis was used to evaluate the association between SS, FIV and mRS. Finally, likelihood ratio test was used to compare model fit between a model including SS alone and model including both SS and FIV. Results: In 332 patients, SS was significantly higher in the poor outcome group (17.3 ± 5.4) when compared to the good outcome group (13.0 ± 6.1) (p=0.0002). In the same analysis, FIVs were also larger in the poor outcome group (110.3 ± 113 cm3) when compared to the good outcome group (37.2 ± 68.3 cm3) (p<0.0001). A combined SS and FIV model correlated significantly better with discharge outcome than did SS alone (p=0.0015). Analysis of 182 patient outcomes at 90 days maintained similar findings, with SS (18 ± 5.9) and FIVs (115.4 ± 121.0 cm3) significantly higher in poor outcomes than in good outcomes; (13.0 ± 5.4) and (35.7 ± 38.2 cm3) respectively (p<0.0001). Combined SS and FIV model, again, was significantly better at modeling outcome at 90 days than was a model including SS alone (p=0.0044). Conclusions: A combined model including FIV and SS better correlates with clinical outcomes at discharge and 90 days in patients with AIS due to LVO, than does a model using SS alone.