Raw data.

Gait prediction is critical in optimizing rehabilitation strategies for stroke survivors. This study evaluates the prognostic utility of the Stroke Rehabilitation Assessment of Movement (STREAM) scores, recorded at admission, for predicting walking ability at discharge and one-month follow-up. We assessed 47 stroke patients using STREAM at admission; walking independence was defined using two criteria: a Functional Ambulation Category (FAC) score >  3 and a 10-Meter Walk Test (10-MWT) speed ≥  0.4 m/s. The predictive validity of STREAM scores was analyzed using the area under the receiver operating characteristic curve (AUC). Sensitivity, specificity, and cut-off values were computed. The analysis revealed that a STREAM score above 38 at admission significantly predicted independent gait by discharge, evidenced by a high AUC of 0.897. At the one-month follow-up, a cut-off score of 29 continued to predict walking independence, with an AUC of 0.987. The subscores further enhanced predictive accuracy and highlighted the effectiveness of the STREAM assessment as a robust predictor of independent walking in stroke patients. These findings suggest the practicality of using STREAM scores to predict walking independence, which can guide the planning of more effective rehabilitation interventions. Trial registration TCTR20240323004 at www.thaiclinicaltrials.org.