Overview

Intravenous thrombolysis with recombinant tissue plasminogen activator (IV tPA; alteplase) improves functional outcomes for selected patients with acute ischemic stroke when given within an appropriate treatment window. The main treatment-related complication that clinicians weigh in shared decision-making is symptomatic intracerebral hemorrhage (sICH): parenchymal bleeding after thrombolysis that is associated with neurological worsening by standardized criteria. The SEDAN score is a simple bedside mnemonic that aggregates five pretreatment factors that repeatedly associate with higher sICH rates in thrombolysis cohorts. It is intended as a risk communication and triage aid, not as a stand-alone rule for eligibility, contraindication, or substitution for guideline-based stroke pathways.

What the score estimates

The SEDAN score estimates the probability gradient of sICH after IV thrombolysis using variables available at the time of treatment decision: serum glucose, noncontrast computed tomography (NCCT) features, age, and stroke severity on the NIH Stroke Scale (NIHSS). The outcome definition used in major validation work has often followed ECASS II–type sICH criteria (substantial parenchymal hemorrhage with clinical deterioration), which differs from other hemorrhage definitions such as those used in safety monitoring studies; reported event rates therefore shift depending on the definition applied to the same dataset.

In very large registry analyses, overall discrimination has been moderate (receiver operating characteristic area under the curve commonly near the mid‑0.60s for ECASS II–defined sICH in broad thrombolysis populations). That performance profile implies the score is helpful for stratifying groups rather than precisely predicting individual outcomes. It should be interpreted beside imaging quality, time from onset, blood pressure trajectory, antithrombotic use, renal function, and institutional treatment protocols.

Mnemonic structure and scoring rules

Each letter in SEDAN corresponds to one binary risk feature. If the feature is present, assign 1 point; if absent, assign 0 points. The total score ranges from 0 (lowest modeled risk in aggregate analyses) to 5 (highest modeled risk).

S — Sugar (admission hyperglycemia)

Hyperglycemia at stroke presentation is common and may reflect stress response, undiagnosed diabetes, or pre‑existing glycemic dysregulation. In thrombolysis risk models, a serum glucose threshold in the range of ≥150 mg/dL (approximately ≥8.3 mmol/L) is typically used as the positive criterion. Mechanistically, higher glucose may relate to blood–brain barrier vulnerability, reperfusion injury, and hemorrhagic transformation, although the score itself does not require users to assume a single causal pathway—only that the variable carries incremental risk in population data.

E — Early infarct signs on pretreatment NCCT

Early ischemic changes on NCCT include loss of gray–white differentiation, hypoattenuation, sulcal effacement, and regional swelling. Many stroke systems operationalize extensive early change using the Alberta Stroke Program Early CT Score (ASPECTS), with lower scores (for example, <7) indicating more established ischemic injury visible before reperfusion therapy. The presence of clearly visible early infarct signs is associated with larger ischemic burden and higher hemorrhagic transformation risk after reperfusion. Local neuroimaging standards should govern how “early change” is adjudicated when the binary item is applied at the bedside.

D — Dense artery sign on pretreatment NCCT

A hyperdense artery sign—most classically hyperdensity in the middle cerebral artery (MCA) distribution—suggests intraluminal thrombus or high protein/concentrated blood products. This finding marks a proximal or substantial arterial occlusion pattern and often correlates with larger infarct territory and more complex reperfusion dynamics. The SEDAN item is scored as present when a relevant dense intracranial arterial sign is identified on the pretreatment scan used for thrombolysis decision-making.

A — Age

Older age is associated with higher sICH rates after thrombolysis in multiple observational datasets. The SEDAN component typically assigns a point when age exceeds a threshold commonly cited as >75 years. Age captures a composite of vascular fragility, microangiopathy burden, and comorbidity, even though chronological age alone does not fully represent individual biological risk.

N — NIH Stroke Scale severity

The NIHSS quantifies stroke severity through standardized examination items spanning level of consciousness, gaze, visual fields, motor function, sensation, language, and neglect. Higher scores reflect more disabling deficits and often larger ischemic involvement. The SEDAN score assigns the NIHSS component when severity exceeds a threshold commonly operationalized as NIHSS >10 (i.e., 11 or higher on the integer scale). This threshold separates milder deficits from moderate-to-severe stroke syndromes in many risk-stratification contexts, though local training and inter-rater reliability still matter for consistent scoring.

How to apply the result in practice

A lower total score is consistent with a lower average sICH frequency in large thrombolysis populations, whereas higher scores identify groups in whom average sICH frequency rises stepwise with each additional point. External validation in a major international thrombolysis registry reported that modeled sICH risk increased with each point increment, with aggregate frequencies on the order of roughly 1–2% at the lowest score stratum versus roughly high teens percent at the maximum score under an ECASS II–type sICH definition—illustrating a wide separation between extremes even though mid-range scores remain clinically common and require individualized judgment.

Practical uses include informed consent discussions (framing bleeding risk alongside expected benefit), periprocedural monitoring intensity (nursing frequency, blood pressure goals, neuro checks), and coordination with neuroimaging follow-up when clinical change occurs. The score should not be used to label thrombolysis as “unsafe” in isolation: many patients with higher SEDAN profiles may still meet guideline criteria and may still derive net benefit depending on onset time, penumbral physiology, treatable occlusion, and alternative options such as endovascular therapy where indicated.

Limitations and common pitfalls

• Definition dependence: sICH rates and model calibration change when switching between ECASS II, SITS-MOST, or radiographic-only hemorrhage definitions.
• Discrimination limits: moderate AUC values mean substantial overlap between patients who do and do not experience sICH; the score is not a diagnostic test with high specificity.
• CT subjectivity: early infarct changes and dense artery signs depend on scan quality, windowing, reader experience, and early time windows where CT sensitivity is limited.
• Evolving stroke care: modern imaging with CT perfusion/MRI, expanded treatment windows, and thrombolysis–endovascular combinations mean risk must be contextualized to contemporary pathways.
• Missing data: the score assumes reliable glucose measurement, accurate NIHSS performance, and timely pretreatment imaging—any gap can misclassify risk.

Documentation and medicolegal considerations

When the SEDAN score is used clinically, document the specific inputs (glucose value and units, NIHSS total, CT impressions for early change and dense sign, age) and clarify that the score is an adjunctive risk estimate rather than a replacement for guideline criteria, institutional protocols, or specialist neuroimaging review. This practice supports clear communication among emergency physicians, stroke neurologists, radiologists, nursing staff, and patients or surrogates during time-sensitive decisions.