Abstract Number: Esoc2026ot192 a Brief Neuropsychological Battery for Detecting Cognitive Impairment Instroke: Validation of the Georgian Ninds-CSN 15-Minute Protocol

Abstract Background and aims Poststroke cognitive impairment (PSCI) is common and adversely impacts recovery and quality of life. While comprehensive neuropsychological assessments are ideal, their time demands limit feasibility in routine care. Methods This study aimed to validate the Georgian adaptation of the NINDS-CSN 15-minute neuropsychological battery based on AHA/ASA and VICCCS-2 consensus recommendations. Results We assessed 73 stroke patients and 102 healthy controls using the NINDS-CSN 30-minute battery, which included Semantic Fluency (Animal Naming), Phonemic Fluency (COWAT), Digit Symbol-Coding (WAIS), and the Hopkins Verbal Learning Test. Supplemental tests included the Georgian versions of the Montreal Cognitive Assessment (MoCA) and Trail Making Tests A and B. The abbreviated 15-minute battery consisted of Semantic Fluency (Animal Naming), Digit Symbol-Coding, Hopkins Verbal Learning Test – Immediate Recall, and Trail Making Tests A and B. Cognitive impairment on the 15-minute battery was defined as performance ≥1.0 standard deviation (SD) below the normative mean in at least two cognitive domains or ≥1.5 SD below the mean in at least one domain, in accordance with harmonization standards. Conclusions The Georgian-adapted NINDS-CSN 15-minute battery demonstrated strong discriminative validity for PSCI, identifying 48 cases with an AUC of 0.8612. For comparison, the 30-minute version identified 46 cases and showed similar diagnostic performance (AUC = 0.86). The slightly higher detection rate may reflect the more inclusive diagnostic threshold. The 15-minute battery achieved 88.9% sensitivity (95% CI: 73.9–96.9%) and 64.5% specificity (95% CI: 45.4–80.8%). Conflict of interest