Figure 2. Effect of Treatment on the Primary Outcome in Subgroups.

The odds ratio in each subgroup was adjusted for the effects of the key variables (age, sex, initial stroke severity, and presence or absence of atrial fibrillation at baseline). Initial stroke severity was measured by means of the National Institutes of Health Stroke Scale (NIHSS; range, 0 to 42, with higher scores indicating more severe neurologic impairment; scores of ≤6 indicate mild impairment, and scores of ≥25 indicate very severe impairment). Age and NIHSS score were used as continuous variables. The variables chosen for multivariate and subgroup analyses were prespecified. The cutoff points for age, NIHSS score, and blood pressure were determined with the use of the Youden method, after a receiver-operating-characteristic analysis tabulating sensitivity, specificity, and accuracy for each possible cutoff point. For time to treatment, patients who did not receive the assigned treatment were excluded, and an additional patient in the endovascular-treatment group was excluded because the procedure was interrupted owing to equipment breakdown. For stroke cause and territory, the 2 patients with conditions mimicking stroke were excluded. For stroke territory, a third patient was excluded because the stroke was in both the posterior and anterior circulation. The analysis of major protocol deviations, and all the other subgroup analyses, included the 12 patients at the center that was withdrawn from the study.