Prediction of long-term outcome by percent improvement after the first day of thrombolytic treatment in stroke patients.

BACKGROUND: We investigated a method for assessing early improvement and predictive factors of early and late outcomes in patients receiving thrombolytic therapy. METHODS: A total of 160 consecutive patients who received thrombolytic therapy were included in the study. Using National Institutes of Health Stroke Scale (NIHSS) scores, percent improvement [(baseline NIHSS score-24-hour NIHSS score)/baseline NIHSS score x 100] was calculated and compared with delta (baseline NIHSS score-24-hour NIHSS score) and with major neurological improvement (MNI, NIHSS score of 0-1 or >or=8 point improvement at 24 h) by receiver operating characteristic (ROC) curve analysis. Finally, we investigated the independent predictors of improvement at 24 h after the thrombolytic therapy and of favorable 3-month outcome (modified Rankin scale score 0-2). RESULTS: By pairwise comparison of ROC curves, percent improvement was stronger than delta (p=0.004) and MNI (p