Takeaway
An electroencephalographic seizure (ES) prediction model could target continuous EEG (CEEG) monitoring to critically ill children at the highest risk for ES, ensuring access to this resource-intense procedure.
Model implementation would reduce the number of patients undergoing CEEG by 29%, however 8% of children experiencing ES would not be identified.
Why this matters
ESs are common (10–40%) in children with acute encephalopathy, but identification of ES requires CEEG monitoring and is resource-intense.
A model employing readily available clinical and EEG variables could target CEEG monitoring to children at highest risk for ES, making it a more viable neuroprotective strategy.