Comparing the Prognostic Impact of Age and Baseline National Institutes of Health Stroke Scale in Acute Stroke due to Large Vessel Occlusion.

Background and Purpose: Little is known about the combined effect of age and National Institutes of Health Stroke Scale (NIHSS) in endovascular treatment (EVT) for acute ischemic stroke due to large vessel occlusion, and it is not clear how the effects of baseline age and NIHSS on outcome compare to each other. The previously described Stroke Prognostication Using Age and NIHSS (SPAN) index adds up NIHSS and age to a 1:1 combined prognostic index. We added a weighting factor to the NIHSS/age SPAN index to compare the relative prognostic impact of NIHSS and age and assessed EVT effect based on weighted age and NIHSS. Methods: We performed adjusted logistic regression with good outcome (90-day modified Rankin Scale score 0­2) as primary outcome. From this model, the coefficients for NIHSS and age were obtained. The ratio between the NIHSS and age coefficients was calculated to determine a weighted SPAN index. We obtained adjusted effect size estimates for EVT in patient subgroups defined by weighted SPAN increments of 3, to evaluate potential changes in treatment effect. Results: We included 1750/1766 patients from the HERMES collaboration (Highly Effective Reperfusion Using Multiple Endovascular Devices) with available age and NIHSS data. Median NIHSS was 17 (interquartile range, 13­21), and median age was 68 (interquartile range, 57­76). Good outcome was achieved by 682/1743 (39%) patients. The NIHSS/age effect coefficient ratio was ([−0.0032]/[−0.111])=3.4, which was rounded to 3, resulting in a weighted SPAN index defined as ([3×NIHSS]+age). Cumulative EVT effect size estimates across weighted SPAN subgroups consistently favored EVT, with a number needed to treat ranging from 5.3 to 8.7. Conclusions: The impact on chance of good outcome of a 1-point increase in NIHSS roughly corresponded to a 3-year increase in patient age. EVT was beneficial across all weighted age/NIHSS subgroups.