Abstract TP096: Brain Health Outcome Scale: A Composite Functional and Cognitive Measure for Post-Stroke Assessment

Background: Stroke survivors frequently experience both functional and cognitive impairments, each contributing substantially to long-term disability, dependence, and healthcare costs. Widely used measures such as the modified Rankin Scale (mRS) capture global motor function but inadequately reflect cognitive status, which affects up to half of survivors and independently predicts adverse outcomes. An integrated approach is needed to more comprehensively assess post-stroke brain health. Methods: We analyzed data from a prospective stroke registry of two hospitals including 1,678 patients assessed with the mRS, the Korean Mini-Mental State Examination (K-MMSE), and the Korean Instrumental Activities of Daily Living (K-IADL). K-MMSE scores were standardized to z -scores using age- and education-adjusted norms, and classified into four cognitive groups: normal ( z ≥ −1), mild (−2 ≤ z < −1), moderate (−3 ≤ z < −2), and severe ( z < −3) cognitive deficit. Six mRS categories (0–5) were combined with these four cognitive groups to form a 24-cell matrix. Candidate composite groupings were generated using a polyomino-based adjacency framework, yielding 408,500 unique configurations. Welch’s F -statistic was applied to maximize between-group differences and minimize within-group variability in K-IADL scores. The top configurations were reviewed for clinical interpretability and ordinal consistency. Results: From the 408,500 candidate configurations, the 20 with the highest Welch’s F -statistic values were selected. The optimal grouping schemes demonstrated clear differentiation in K-IADL scores, with no overlap in the corresponding 95% confidence intervals. Moreover, the selected configurations exhibited a consistent sequential ordering of groups from the top-left to the bottom-right of the matrix, enhancing both interpretability and practical applicability. Conclusions: The BHOS offers a novel, integrated measure of post-stroke brain health that differentiates recovery profiles more effectively than existing scales. Its clear ordinal structure facilitates interpretation and may enhance prognostic accuracy, guide individualized rehabilitation, and serve as a robust endpoint in clinical trials. Future studies will assess its predictive validity for long-term outcomes, including mortality, institutionalization, and dementia.