Association of emergency department stroke care processes with rehabilitation outcomes after first-ever stroke

To examine the association between emergency department (ED) processes and acute interventions after first-ever stroke and the functional and motor gains achieved by the end of a rehabilitation program in a physical medicine and rehabilitation clinic.

This single-center retrospective observational cohort study included 257 patients who presented to the ED with a first-ever stroke and subsequently underwent a rehabilitation program in a physical medicine and rehabilitation clinic. Demographic and clinical characteristics, ED process/intervention variables [large vessel occlusion (LVO), stroke code activation, intensive care unit admission, intravenous thrombolysis (IVT), endovascular therapy (EVT), initiation of antithrombotic therapy, and time metrics], and rehabilitation outcomes were extracted from electronic records. Functional and motor outcomes were assessed using the Functional Independence Measure (FIM), Functional Ambulation Scale (FAS), and Brunnstrom staging. The primary outcome was clinically meaningful functional improvement (ΔFIM ≥ 22). Secondary outcomes included ΔFIM, FIM efficiency, ΔFAS ≥ 1, independent ambulation at the end of the rehabilitation episode (FAS ≥ 4), and ≥ 1-stage gain in Brunnstrom staging for the upper extremity, hand, and lower extremity. Multivariable logistic regression and receiver operating characteristic (ROC) analyses were performed.

The mean age was 65.02 ± 13.55 years, and 38.9% of patients were female. Significant improvement was observed across functional and motor outcomes during rehabilitation. Total FIM increased from 75.65 ± 15.90 to 96.42 ± 17.02 (ΔFIM 20.77 ± 13.70; p < 0.001; Cohen's dz = 1.52), median FAS increased from 4 to 5 (p < 0.001), and 78.6% of patients achieved independent ambulation at the end of the rehabilitation episode (FAS ≥ 4). Clinically meaningful functional improvement (ΔFIM ≥ 22) was observed in 45.9% of patients. Brunnstrom stages improved significantly in all domains (all p < 0.001), with ≥ 1-stage gain rates of 58.6% for the upper extremity, 55.5% for the hand, and 64.5% for the lower extremity. In multivariable analyses, stroke code activation and IVT were associated with several rehabilitation outcomes, whereas LVO was associated with less favorable outcomes in selected domains. For clinically meaningful functional improvement, female sex, stroke code activation, IVT, and initiation of antithrombotic therapy were positively associated, whereas NIHSS, baseline FIM, and LVO were negatively associated. ROC analysis showed modest discrimination for the ΔFIM ≥ 22 model (AUC = 0.688) and good discrimination for the FAS ≥ 4 model (AUC = 0.878).

Emergency department organization and acute stroke care processes may be associated not only with early clinical outcomes but also with functional and motor gains achieved during rehabilitation. In particular, stroke code activation and IVT were associated with more favorable rehabilitation outcomes in several models, whereas LVO was associated with less favorable gains in selected domains. These findings are consistent with the value of integrated stroke care pathways linking emergency care and rehabilitation.