Abstract 3436: Specific NIH Stroke Scale Items Strongly Predict Hospital Arrival Mode, Thrombolysis Administration, and Clinical Outcomes in Acute Ischemic Stroke

Abstract

Background: The 11-item National Institutes of Health Stroke Scale (NIHSS) is widely used as an index of stroke severity and prognostication. However, no studies have specifically examined the influence of NIHSS items on care processes and outcomes in Acute Ischemic Stroke (AIS). Furthermore, potential distinctions in neurologic signs of AIS that may contribute to disparities in race-ethnic treatment rates and outcomes have not been evaluated. We assessed the relation of neurological signs on the NIHSS to arrival mode, thrombolysis treatment and clinical outcomes in AIS, and also evaluated the influence of race-ethnicity. Methods: We analyzed the dataset of a hospital network comprising prospectively collected data on AIS patients presenting within 12 hours of ictus between June 2004 and May 2011. Outcomes evaluated were mode of arrival (ambulance vs. other), IV thrombolysis (yes vs. no), discharge destination (home vs. other), unfavorable day-90 functional activity (modified Rankin Scale (mRS) score >1), unfavorable day-90 disability (Barthel Index <95), and day-90 mortality. Outcomes were adjusted for pre-specified covariates in a multivariable logistic regression model. Results: Of 972 AIS patients 462 (48%) were women, 635 (65%) Non-Hispanic White, 162 (17%) White Hispanic, 106 (11%) Black, and 69 (7%) other race/ethnicity. Overall, the presence of extinction/neglect was the strongest predictor of arriving by ambulance (adjusted OR 2.32, 95% CI: 1.53-3.51), and abnormal level of consciousness (LOC) was the strongest predictor of receipt of IV thrombolysis (adjusted OR 2.25, 95% CI: 1.67-3.04), while limb ataxia was the only NIHSS item not significantly associated with either arrival mode or thrombolysis treatment. Presence of gaze preference was the strongest predictor of not going home directly from the hospital (adjusted OR 0.2, 95% CI: 0.14-0.29), unfavorable day-90 functional activity (adjusted OR 0.21, 95% CI: 0.12-0.37) and poor mortality outcome (adjusted OR 5.92, 95% CI: 3.42-10.25), while abnormal LOC was the strongest predictor of unfavorable day-90 disability (adjusted OR 0.27, 95% CI: 0.15-0.47). White Hispanic AIS patients with sensory symptoms were less likely to arrive by ambulance (adjusted OR 0.31, 95%CI: 0.13-0.74) but more likely to go home directly (adjusted OR 2.81, 95% CI: 21.31-6.02), while Black AIS patients with abnormal level of consciousness were more likely to receive IV thrombolysis (adjusted OR 4.69, 1.80-12.26). Conclusions: Specific items on the NIHSS are strongly related to hospital arrival mode, thrombolysis treatment, and clinical outcomes among AIS patients. Some of these associations vary by race and ethnicity. These results could aid prognostication and identify areas in the community, pre-hospital and emergency department phases of stroke care requiring more education, training, or intervention, to boost AIS outcomes.