MANA nomogram

MANA nomogram estimates the probability of malignant cerebral edema (MCE) in patients with large hemispheric infarction (LHI) using four clinical and imaging predictors.

Key Features:

• Predictors: Previously known atrial fibrillation (KAF), midline shift (MLS), National Institutes of Health Stroke Scale (NIHSS) score, and anterior cerebral artery (ACA) territory involvement.
• Modeling approach: Binary logistic regression was used to identify independent predictors and derive the nomogram.
• Outcome definition: MCE defined as death or requirement for decompressive hemicraniectomy (DHC) due to mass effect indicated by midline shift ≥5 mm or effacement of basal cisterns.
• Dataset: Retrospective cohort of 142 consecutive LHI patients within 24 hours of onset (January 1, 2016 to August 31, 2019).
• Implementation: Dynamic nomogram implemented using the Dynnom package.
• Performance: C-statistic 0.887 ± 0.041 and AUC-ROC 0.887 (95% CI 0.828–0.934).

Scientific Applications:

• Risk stratification: Estimate individual risk of developing MCE after LHI to support clinical risk stratification.
• Therapeutic decision support: Inform consideration of decompressive hemicraniectomy and intensified monitoring for patients at high risk of MCE.

Methodology:

Retrospective analysis of 142 consecutive LHI patients within 24 hours of onset; MCE defined as death or DHC for mass effect (MLS ≥5 mm or basal cistern effacement); binary logistic regression identified KAF, MLS, NIHSS, and ACA involvement as independent predictors, and a dynamic nomogram was implemented using the Dynnom package.