Abstract 3347: Derivation and Double External Validation of an Integer-based Prognostic Score (ASTRAL Score) to Predict Functional Outcome in Acute Ischemic Stroke: a Collaborative Analysis from the ASTRAL, Athens and Vienna Stroke Registries

Abstract

Introduction: We aimed to develop and externally validate a simple, integer-based prognostic score to predict functional outcome at three months in patients with acute ischemic stroke (AIS) using variables readily available at hospital admission. Methods: Using emergently available demographic, clinical and laboratory variables of consecutive AIS patients in a single university hospital, logistic regression analysis was performed in the derivation cohort (ASTRAL registry) to identify independent predictors of unfavorable outcome, defined as modified Rankin Scale score >2. The β-coefficient of each covariate of the fitted multivariate model was used to generate an integer-based point scoring system for each one; the overall score was calculated as the sum of weighted scores of the covariates. The model was validated externally in two AIS cohorts from the Athens and Vienna Stroke Registries. Two indices of accuracy were assessed: discrimination (Area Under the Curve, AUC) and calibration (Hosmer-Lemeshow test). Results: Six covariates were identified as independent predictors of unfavorable outcome in 1968 patients in ASTRAL: age (A), severity of stroke (S) measured by admission NIHSS, time from stroke onset-to-admission (T), range of visual fields on clinical exam (R), acute serum glucose (A) and level of consciousness (L). AUC in the ASTRAL cohort was 0.850. β-coefficients were multiplied by 4 and rounded to the closest integer. The score was well calibrated in the derivation (p=0.43) and validation cohorts [0.22 and 0.49 in the Athens (n=1659) and Vienna cohorts (n=653) respectively, 0.82 in their combination]. AUCs were 0.937 and 0.771 in the Athens and Vienna cohorts respectively, and 0.902 in their combination. An ASTRAL score of 28 indicates a 50% likelihood of unfavorable outcome. Conclusions: The ASTRAL score is a simple integer-based score to predict 3 months functional outcome using 6 readily available items at hospital admission. It performed well in double external validation and may be a useful tool for clinical practice and stroke research.