ABSTRACT NUMBER: ESOC2026A1786 ANALYSIS OF CHARACTERISTICS AND CLINICAL OUTCOMES AFTER ISCHAEMIC STROKE BY NIHSS IN A NATIONALLY REPRESENTATIVE US ELECTRONIC HEALTH RECORDS DATABASE

Abstract Background and aims NIHSS-defined stroke severity strongly influences outcomes and treatment decisions after IS. We assessed real-world patient characteristics and clinical outcomes according to baseline NIHSS in a nationally representative US-based cohort. Methods In this retrospective observational study, adults hospitalised between 01/2017–12/2023 with a first IS, without a known cardioembolic source at admission, were identified from the Truveta EHR database. NIHSS scores were calculated from structured data and clinical notes processed with Truveta’s large language model. Baseline patient characteristics outcomes (mortality, recurrence, haemorrhage) up to 1-year were evaluated. Results Of 180,021 patients with IS, 67% (n=120,923) had estimable admission NIHSS scores (≤5: 78%; 6–15: 14%; >15: 8%). Relatively few patients had NIHSS >15 (n=9636). Patients with NIHSS >15 were older (mean age: 70.4 years) and more frequently had LAA, cardioembolic, or combination TOAST subtypes than those with NIHSS ≤5. The risk of recurrent IS, all-cause mortality, composite of all-cause mortality, MI or stroke, as well as ISTH major bleeding, and symptomatic ICH increased with higher NIHSS score categories across all time points up to 1 year (Figure). Almost 50% of patients with NIHSS >15 experienced the composite outcome of stroke, all-cause mortality or MI at 1 year compared with only 17% of patients with NIHSS ≤5. Conclusions Higher initial NIHSS scores correlate with worse clinical outcomes throughout the 1-year follow-up, highlighting an elevated risk of poor prognosis. Conflict of interest This study was funded by Bayer AG. MR declares no conflict of interest. TN reports serving as an Associate Editor of Stroke and consulting relationships with Bayer, Medtronic and Route92. SS’s institution received compensation from Bayer. AS reports research grants/contracts from AstraZeneca, Bayer, the Canadian Institutes of Health Research, Daiichi Sankyo Company, the Heart and Stroke Foundation of Canada, the National Institutes of Health, Octapharma USA, Inc. and Servier Affaires Medicales; has served on advisory boards for AstraZeneca, Bayer, Daiichi Sankyo Company and Takeda Pharmaceutical Company; and has served on a data and safety monitoring board for Bayer. MD and NT are employees of Truveta. SK, DRM, JX and KA are employees of Bayer. Figure 1 - belongs to Conclusions