Abstract Details

Title Identifying Key Predictors of Stroke Recurrence and Survival using a Propensity Score Matched Analysis

Topic Cerebrovascular Disease and Interventional Neurology

Presentation(s) P4 - Poster Session 4 (5:00 PM-6:00 PM)

Poster/Presentation
Number 14-011

Objective This study uses propensity score matching (PSM) to reduce selection and confounding bias and aims to accurately identify predictors of stroke recurrence and 1-year survival.

Background Stroke recurrence poses a significant challenge to the management of long-term outcomes in stroke survivors. Traditional studies are often limited by residual bias and a narrow focus on isolated risk factors, potentially obscuring the complex interplay between clinical outcomes and patient characteristics.

Design/Methods We retrospectively analyzed 39,947 ischemic stroke patients from the Regional Center Stroke Registry in Korea. Using PSM, patients were stratified into categories (>5, >16, >21) according to their discharge NIHSS score to ensure balanced comparisons. Logistic regression was used to assess the effect of NIHSS scores and other covariates on the probability of one-year recurrence, while Kaplan-Meier estimates were used to assess survival probabilities.

Results The incidence of early ischemic stroke recurrence was 3.83% (n=1,531). Higher discharge NIHSS scores significantly correlated with increased risk of recurrence within one year. Specifically, patients with scores above 16 and especially above 21 had the highest risk. Kaplan-Meier analysis showed a significant decrease in survival with higher NIHSS thresholds, underscoring their prognostic importance.

Conclusions Discharge NIHSS score is an important and reliable predictor of stroke recurrence and survival, confirming its utility in post-stroke prognosis. The use of PSM has improved the reliability of these findings, highlighting the importance of NIHSS for risk stratification and management in stroke care.

Authors/Disclosures
Jung Keun Hyun, MD, PhD PRESENTER	Prof. Hyun has nothing to disclose.
Yuna Kim (Dankook University)	No disclosure on file

��ɫ��

��ɫ��