Introduction Early prediction of long-term recovery after endovascular thrombectomy (EVT) is essential for guiding care but is often confounded by the instability of acute neurological assessments. We aimed to evaluate the prognostic value of the National Institutes of Health Stroke Scale (NIHSS) measured at day 5–7 or earlier if discharged (D5–7/discharge) compared with baseline NIHSS and change-based metrics for predicting 90-day functional outcomes. Methods We performed a post hoc analysis of the MARVEL trial, a multicenter randomized clinical trial involving adult patients with acute ischemic stroke due to anterior circulation large-vessel occlusion. The prognostic performance of baseline NIHSS, D5–7/discharge NIHSS, absolute change (ΔNIHSS), and percentage change was evaluated. The primary outcome was favorable outcome, defined as a modified Rankin Scale (mRS) score of 0–2 at 90 days. Discriminative performance was assessed using the area under the receiver operating characteristic curve (AUC), net reclassification improvement (NRI), and integrated discrimination improvement (IDI). Results Among 1650 included patients, the D5–7/discharge NIHSS showed the highest discriminative ability among the evaluated NIHSS-based measures for predicting favorable outcome (AUC 0.897; 95% confidence interval CI 0.881–0.912), exceeding percentage change (AUC 0.891), ΔNIHSS (AUC 0.866), and baseline NIHSS (AUC 0.650) (all p < 0.001). A pragmatic, restricted cubic spline-informed threshold of D5–7/discharge NIHSS ≤ 5 yielded a sensitivity of 64.7% and specificity of 93.3% for 90-day favorable outcome. Adding D5–7/discharge to a model containing traditional covariates significantly improved predictive accuracy (AUC increased from 0.782 to 0.910; NRI 0.437; p < 0.001). Percentage change in NIHSS offered better discrimination than absolute change. Conclusion The NIHSS assessed at D5–7/discharge showed higher discriminative performance for 90-day functional outcomes than baseline and change-based measures in this post hoc analysis. A threshold of ≤ 5 provided high specificity and may aid bedside counseling and discharge planning, although external validation is warranted. Trial Registration ChiCTR.org.cn Identifier, ChiCTR2100051729.
Chen et al. (Sat,) studied this question.