Electrographic Seizure Characteristics and Electrographic Status Epilepticus Prediction

Area of Science:

• Pediatric Neurology
• Neurocritical Care
• Epileptology

Background:

• Critically ill children with encephalopathy often undergo continuous EEG monitoring (CEEG) to detect non-convulsive seizures.
• Electrographic seizures (ES) and electrographic status epilepticus (ESE) are critical findings, but their characteristics and optimal prediction strategies require further elucidation.

Purpose of the Study:

• To characterize electrographic seizures (ES) and electrographic status epilepticus (ESE) in critically ill children.
• To evaluate a predictive model for ESE to guide continuous EEG monitoring (CEEG) utilization.

Main Methods:

• Prospective observational study of 1,399 critically ill children with encephalopathy undergoing CEEG.
• Descriptive statistics to characterize ES and ESE.
• Development and evaluation of a model for ESE prediction.

Main Results:

• ES occurred in 25% of subjects; 23% of those had ESE.
• ESE most commonly presented as recurrent brief seizures (63%) rather than prolonged individual seizures (37%).
• An ESE prediction model showed an AUC of 0.81, but with 91% sensitivity, specificity was 56%, leading to a 11% PPV and 99% NPV given a 6% ESE incidence.

Conclusions:

• ESE is frequent in critically ill children, often manifesting as recurrent brief seizures.
• An ESE-exclusive prediction model may offer limited improvement in CEEG utilization and risks missing ESE cases.
• Models identifying risk for any ES may be more beneficial for preventing progression to ESE.