A simple risk index predicts endovascular treatment outcomes in acute ischemic stroke: prognostic value of SPAN-100

You're not even trying for 100% recovery! You're fired!

“What's measured, improves.” So said management legend and author Peter F. Drucker 

 A simple risk index predicts endovascular treatment outcomes in acute ischemic stroke: prognostic value of SPAN-100

Noémie Ligot¹Marie Dagonnier²Boris Lubicz³Nicolas Brassart⁴Gilles Naeije^1*

• ¹Department of Neurology, Hôpital Erasme, Université Libre de Bruxelles, Brussels, Belgium
• ²Department of Neurology, CHU HELORA—Kennedy, Mons, Belgium
• ³Department of Interventional Neuroradiology, Hôpital Erasme, Université Libre de Bruxelles, Brussels, Belgium
• ⁴Department of Interventional Vascular Radiology, CHU HELORA—Kennedy, Mons, Belgium

Background: Mechanical thrombectomy (MT) is a proven intervention for patients with acute ischemic stroke (AIS) due to a large vessel occlusion (LVO). However, outcomes after MT remain variable, particularly in high-risk groups. The SPAN-100 index, a simple bedside tool combining age and baseline NIHSS scores, has been associated with poor outcomes in AIS patients treated with intravenous thrombolysis. This study evaluates the prognostic value of the SPAN-100 index and its weighted variant (wSPAN) in predicting outcomes following MT.

Methods: We conducted a retrospective cohort study of patients with AIS who were treated with MT between 2015 and 2024 at two Belgian university hospitals. SPAN and wSPAN scores were calculated at admission, and patients were stratified by SPAN ≥100 vs. <100. The primary outcome was functional status at 90 days, as assessed using the modified Rankin Scale (mRS). Secondary outcomes included mortality (mRS = 6) and favorable outcomes (mRS ≤ 3). Discriminative ability was assessed using receiver operating characteristic (ROC) analysis (AUC, 95% CI), and model performance was evaluated using AIC and BIC. The positive predictive value (PPV) and negative predictive value (NPV) were calculated for SPAN ≥100, and outcome proportions were compared between SPAN-defined groups using Fisher’s exact test.

Results: A total of 530 patients were included, of whom 116 had SPAN scores of ≥100. These patients experienced significantly worse outcomes, with higher mortality (60.0% vs. 17.6%) and lower rates of mRS ≤ 3 (19.2% vs. 71.6%, both p-values < 0.001). However, among survivors with SPAN scores of ≥100, nearly half of them achieved recovery(So failed at 100% recovery!)(mRS ≤ 3). ROC analysis showed good discrimination for both SPAN and wSPAN: AUCs were 0.77 and 0.78 for mRS ≤ 3 and 0.80 and 0.81 for mortality. wSPAN showed a slightly better model fit (lower AIC/BIC). The SPAN ≥100 threshold had a high PPV for poor outcomes and a high NPV for survival.

Conclusion: SPAN-100 and wSPAN are pragmatic and reliable prognostic tools for AIS patients undergoing MT. While a SPAN score ≥100 identifies a high-risk group with poorer overall outcomes, it should not be used as an exclusion criterion. Many patients with a SPAN score ≥100 achieved functional recovery(So failed at 100% recovery!), supporting MT as a justified intervention even for older, more severely affected individuals. These findings highlight the importance of combining risk stratification with clinical judgment rather than relying on rigid thresholds.

1 Introduction

Mechanical thrombectomy (MT) has revolutionized the treatment of acute ischemic stroke (AIS) caused by large vessel occlusion (LVO), offering substantial improvements in functional outcomes when performed promptly. Despite the rapid evolution of mechanical thrombectomy (MT) as the standard of care for acute ischemic stroke due to large vessel occlusion, outcome prediction remains challenging. Numerous multivariable models have been proposed, incorporating clinical variables (1, 2), advanced neuroimaging (core infarct volume, ASPECT, collateral status, MRI-based penumbra) (3–6), or procedural metrics (e.g., time to recanalization) (5, 7). However, even when combining these features, the added prognostic value remains limited, and model performance is often suboptimal in external validation (1, 2). More importantly, such models are frequently complex, require unavailable or delayed inputs, and lack bedside applicability. This complexity underscores the need for reliable, practical prognostic tools to assist clinicians in making informed treatment decisions and to guide discussions with patients and their families about expected outcomes in these high-stakes scenarios (8, 9).

The Stroke Prognostication using the age and NIH Stroke scale (SPAN-100) index is a simple and pragmatic tool that sums age and baseline National Institutes of Health Stroke Scale (NIHSS) score to estimate stroke outcomes. Patients with SPAN-100 scores of 100 or greater (SPAN≥100) are typically older with more severe neurological deficits, characteristics that have been consistently associated with poorer prognoses. Studies in patients treated with intravenous thrombolysis (IVT) have demonstrated that SPAN ≥100 status predicts higher rates of mortality and severe disability at 90 days (10, 11). However, the applicability of the SPAN-100 index in patients undergoing MT remains seldom underexplored (12). This knowledge gap is particularly important (13) given that MT is a highly effective intervention for LVO, even in some high-risk populations (14) and when subgroups are weighted for age/NIHSS, despite the combined adverse effect of age and NIHSS on MT outcomes (13).

Understanding the role of SPAN-100 in MT could refine prognostication and enhance decision-making by helping clinicians balance the risks and benefits of treatment for patients with complex profiles. Furthermore, prognostic tools such as SPAN-100 provide a valuable framework for counseling patients and their families about expected functional outcomes, fostering shared decision-making and realistic goal setting in the post-stroke recovery process. By identifying high-risk patients who might still benefit from MT, the SPAN-100 index could optimize resource allocation and improve individualized care strategies in AIS management.

This study aims to evaluate the SPAN-100 index, based solely on age and NIHSS, and its weighted variant (wSPAN, which consists of the sum of age added to three times the NIHSS) (13) as practical bedside tools for prognostication in a real-world multicenter thrombectomy cohort. The SPAN framework is intentionally minimalistic but anchored on the two most influential and universally available predictors of stroke outcome. We hypothesized that SPAN and wSPAN would demonstrate discriminative performance comparable to more elaborate models, thus offering a pragmatic solution for rapid risk stratification.

More at link.