The FAITHS2 Model Predicts Functional Disability in Patients With Acute Ischemic Stroke

Abstract

The present study aimed to develop a model to predict functional disability at 3 months in patients with acute ischemic stroke (AIS) ( n = 5,406). The primary outcome was functional disability (modified Rankin Scale [mRS] >2) at 3 months. A prediction model including blood biomarkers was developed based on a multivariable logistic regression model, which was internally validated by the 100-time bootstrap method. A nomogram and a web-based calculator were developed for usage in clinical practice. At 3 months, 11% (638/5,406) of the patients had functional disability. Seven independent predictors of functional disability at 3 months were incorporated into the FAITHS2 model (fasting plasma glucose, age, interleukin-6, stroke history, National Institute of Health Stroke Scale [NIHSS] at admission, sex, and systolic blood pressure). The Area Under Curves (AUCs) were 0.814 (95% confidence interval [CI] 0.796–0.832) and 0.808 (95% CI 0.806–0.810), and the Brier scores were 0.088 ± 0.214 and 0.089 ± 0.003 for the derivation cohort and internal validation, respectively, showing optimal performance of the model. The FAITHS2 model has excellent potential to be a dependable application for individualized clinical decision making.