Updated 30 Aug 2026
Seizure

SeLECT 2.0 Calculator — Post-Stroke Epilepsy Risk

Severity, Large-artery disease, Early seizure, Cortical involvement, Territory — the validated model for late-seizure risk after ischemic stroke, updated to weight acute symptomatic status epilepticus. What it changes is counseling, not prophylaxis.

Reviewed by Zaka Ahmed, MD Clinical reference · 4 min read

Post-stroke epilepsy prediction

SeLECT 2.0

Score each item for an ischemic stroke (the model does not apply to ICH or SAH). Total 0–13; the acute symptomatic seizure item distinguishes single seizures from status epilepticus per the 2.0 update.

1Severity of stroke SeNot scored
2Large-artery atherosclerosis LNot scored
3Early seizure (≤7 days) ENot scored
4Cortical involvement CNot scored
5Territory of MCA involvement TNot scored
/ 13
5 items remaining

Score every item to see the late-seizure risk band.

What the score does and does not change

    Educational tool only. Risk bands shown are approximations of the published SeLECT estimates (exact percentages per score are in Galovic et al.); the model applies to ischemic stroke only and does not justify prophylactic antiseizure medication at any score. It does not replace clinical judgment.

    Clinical notes & interpretation Scoring guidance, pitfalls, FAQs, and references

    Two different animals: early seizures and post-stroke epilepsy

    Seizures within 7 days of stroke are acute symptomatic — provoked by the acute injury, with low intrinsic recurrence risk once the acute phase passes — and by themselves do not constitute epilepsy or mandate long-term treatment. Seizures after day 7 are unprovoked, carry recurrence risk high enough that a single one effectively diagnoses post-stroke epilepsy, and usually do warrant an antiseizure medication. SeLECT exists to quantify the bridge between the two: given this stroke and this acute course, how likely is that late seizure? The 2.0 revision recognized that an early status epilepticus predicts far more late epilepsy than a single early seizure, weighting it 7 points versus 3.

    What to do with the number

    Nothing pharmacologic — that is the point most worth writing down. No trial supports prophylactic ASMs to prevent post-stroke epilepsy, and guidelines recommend against them; sedating an elderly stroke patient with levetiracetam "just in case" trades real cognitive cost for no proven benefit. The score's outputs are softer but real: risk-proportionate counseling of patient and family (what a focal seizure looks like in an aphasic patient is genuinely non-obvious), driving guidance, a documented plan for what to do about the first suspicious spell, and — at the top of the range — arranging follow-up that includes EEG access rather than a generic six-month visit. It also disciplines the inpatient reflex in the other direction: the patient who seized once at onset and left the hospital on indefinite levetiracetam deserves a deliberate deprescribing conversation, not automatic refills.

    Frequently asked questions.

    Does SeLECT apply to hemorrhagic stroke?

    No — it was derived and validated in ischemic stroke. ICH has its own (higher) epileptogenicity and its own model (CAVE score: cortical location, age <65, volume >10 mL, early seizures).

    Should a high SeLECT score trigger prophylactic ASMs?

    No. There is no evidence that prophylaxis prevents epileptogenesis after stroke, and guidelines recommend against primary prophylaxis. The score changes counseling and follow-up, not prescriptions.

    What about the patient started on levetiracetam after an early seizure?

    Treating through the acute phase is reasonable; continuing indefinitely usually is not. A common approach is to taper after the acute period if the seizure was clearly acute symptomatic and imaging/EEG do not argue otherwise — an individualized decision, and exactly the conversation SeLECT informs.

    References.

    1. Galovic M, Döhler N, Erdélyi-Canavese B, et al. Prediction of late seizures after ischaemic stroke with a novel prognostic model (the SeLECT score): a multivariable prediction model development and validation study. Lancet Neurol. 2018;17(2):143–152. PubMed
    2. Sinka L, Abraira L, Imbach LL, et al. Association of mortality and risk of epilepsy with type of acute symptomatic seizure after ischemic stroke and an updated prognostic model (SeLECT 2.0). JAMA Neurol. 2023;80(6):605–613.
    3. Holtkamp M, Beghi E, Benninger F, et al. European Stroke Organisation guidelines for the management of post-stroke seizures and epilepsy. Eur Stroke J. 2017;2(2):103–115. PubMed