END-IT Scale Shows Only Moderate Accuracy for Status Epilepticus Outcome

By Marcus Bennett

The END-IT score predicted 3-month functional outcome after status epilepticus with only moderate accuracy in an external validation of 114 adults at Cairo University Hospital, according to a prospective cohort study run from August 2024 to August 2025. For clinicians who triage SE patients in emergency departments and intensive care units, the message is blunt: do not treat the score as a standalone verdict.

The study enrolled consecutive adult patients presenting with SE to the emergency department or intensive care units at the Egyptian tertiary-care academic centre. Investigators collected data prospectively throughout hospitalization and followed every patient for three months, scoring functional outcome on the Modified Rankin Scale. An unfavorable outcome meant an mRS of 3–6 — death or functional dependence on others.

What did the cohort look like?

The Egyptian sample differed sharply from the population in which END-IT was originally derived, and that difference sits at the heart of the findings.

• Median age was 33 years (interquartile range 17.75–54.25), older than the 25.5-year median of the original derivation cohort.
• Males made up 63.2% of cases.
• A known history of epilepsy was present in 55.3%.
• Etiologies: epilepsy-related causes 33.3%, vascular disorders 24.6%, CNS inflammation or infection 21.9%, metabolic derangements 16.7%, CNS tumors 2.6%, degenerative disorders 0.9%.

The authors note the profile matches other cohorts from developing countries, where epilepsy-related causes such as non-adherence to antiseizure medications dominate. That contrasts with the derivation setting, where encephalitis appeared in 35.6% of patients and 31.1% needed mechanical ventilation.

Why external validation mattered here

END-IT was, until now, the only scale built explicitly to predict the 3-month adverse outcome of SE rather than in-hospital mortality. Its five items — encephalitis, non-convulsive SE, diazepam resistance, imaging findings, and tracheal intubation — were tuned on a young patient population with a heavy inflammatory burden. Prognostic models often drift when the case mix changes. The Cairo team set out to test exactly that drift, and their conclusion acknowledges it: performance was moderate, not decisive.

The study also compared END-IT against the Status Epilepticus Severity Score (STESS) within the same cohort, and explored clinical variables outside the score's five items as complementary predictors. Both STESS and the Epidemiology-Based Mortality Score in Status Epilepticus (EMSE) were originally designed for in-hospital mortality, not functional outcome — a distinction reading-room neurologists and ICU teams should keep in mind when choosing a bedside tool.

What changes at the bedside?

Three practical takeaways emerge from the data.

• Treat END-IT as a screening aid, not a prognosis. Moderate performance in a real-world, epilepsy-heavy cohort means the score narrows risk; it does not settle it.
• Ask about comorbidity. The authors found that incorporating comorbidity burden may further improve prognostic assessment beyond the five standard items.
• Match the tool to the question. STESS and EMSE answer mortality; END-IT answers 3-month dependence. Conflating the two outputs invites error.

The researchers received no specific grant from public, commercial, or not-for-profit funders, and they declare no competing financial interests. Rehab Magdy led the work, with Nirmeen Kishk, Ehab Shaker, Amani Nawito, Alshimaa S. Othman, and Samar Elsayed Abdelal contributing to supervision, analysis, and data curation.

The authors themselves frame the next step plainly: findings require validation in larger multicenter cohorts before the comorbidity-augmented approach enters routine use. They add that refining SE prognostication will ultimately support more informed decisions about intervention strategies — a goal this Egyptian cohort moves one external-validation step closer to meeting.