Poor Performance of Stroke Prognostication Using Age and National Institutes of Health Stroke Scale-100 to Predict 3- and 12-month Outcomes of Ischemic Stroke in China National Stroke Registry

BackgroundStroke Prognostication using Age and NIH (National Institutes of Health) Stroke Scale (SPAN)-100 is a simple and easy-to-use tool for assessing the outcomes of ischemic stroke after thrombolysis. To explore its application, we evaluated SPAN-100's prognostic value in predicting 3- and 12-month outcomes in general ischemic stroke patients.MethodsWe applied the SPAN-100 to ischemic stroke patients from the China National Stroke Registry. Poor outcome was defined as a modified Rankin Scale of 2-6. Discrimination of SPAN-100 was assessed by the area under the receiver–operator curves (AUC) and 95% confidence intervals (CI). We also performed an exploratory post hoc analysis of the performance of the SPAN index score using 80 as the cutoff point.ResultsAmong 11,894 ischemic stroke patients, 479 (4.0%) patients were SPAN-100 positive. The AUC of SPAN-100 for poor outcome was .54 (95% CI, .54-.54) at 3 months and .54 (95% CI, .54-.55) at 12 months, respectively. In the exploratory analysis, when 80 was used as the cutoff point of SPAN index score, the AUC for poor outcome was .66 (95% CI, .66-.67) at 3 months and .68 (95% CI, .67-.68) at 12 months, respectively.ConclusionsSPAN-100 suffered from low prediction power for 3- and 12-month outcomes of ischemic stroke in Chinese population. A cutoff point of 80 may improve the performance, but none of them had an AUC above the threshold of .8 required for use in individuals.