Background: Ischemic stroke is a leading cause of morbidity. Assessing the chances of recovery is critical to optimize poststroke care.
Methods: We used a cohort of patients from the Virtual International Stroke Trial Archive that participated in acute stroke trials (control arm) and were followed for 90 days. The cohort was grouped by day 7 (D7) modified Rankin scale (mRS) scores. Variables that were associated with good outcome (mRS 0-2 at 90 days) in the univariate analysis were entered into a logistic regression model to determine the independent good outcome criteria for each D7 mRS tier.
Results: We analyzed 1,798 patients. The independent good outcome criteria identified for different mRS tiers were: D7 mRS of 3: age < or =70, 0-2 vascular risk factors, D7 NIH Stroke Scale (NIHSS) arm strength < or =1, D7 NIHSS language score = 0; D7 mRS of 4: age < or =70, male, D7 NIHSS facial palsy < or =1, D7 NIHSS visual = 0, D7 NIHSS leg strength < or =1, D7 NIHSS dysarthria = 0; D7 mRS of 5: age < or =70, IV tPA treatment, D7 NIHSS dysarthria = 0, D7 NIHSS leg strength < or =2. For each mRS tier, we observed a graded increase in the percentage of the primary and secondary end points with increase in the number of criteria.
Conclusions: We identified clinical variables that predict good outcome, are specific to each day 7 mRS tier, and enable easy and informative assessment of the patient's likelihood of achieving varying degrees of recovery at day 90. These results may be useful in both clinical practice and research but require validation in an independent patient cohort.
Copyright 2009 S. Karger AG, Basel.