A new rating system can accurately predict which critically ill patients are in danger of having seizures while hospitalized, a large, multi-national trial shows.

Epilepsy physician and researcher Aaron Struck, MD, is lead author of a study published today in JAMA Neurology that validates the 2HELPS2B algorithm as way of predicting which patients are at greatest risk.

Aaron Struck
Aaron Struck

“There is growing recognition that many of the sickest patients in the hospital are having seizures that can complicate the course of their illness,’’ says Struck, assistant professor of neurology at the University of Wisconsin-Madison School of Medicine and Public Health.

“These types of seizures are not like the typical seizures in patients with epilepsy – where the seizures are readily obvious to anyone nearby. In these sick patients with seizures, there is rarely any obvious outward sign that doctors or nurses could recognize.”

Struck explains that the 2HELPS2B algorithm is a scoring system that identifies patients at greatest risk for seizures. The study looked at continuous electroencephalogram (cEEG) data from more than 2,000 patients who were in five major medical centers: University Hospital in Madison, where Struck practices; Duke University; Medical University of South Carolina; Massachusetts General Hospital, Emory University (Grady Memorial Hospital), and the Free University of Brussels in Belgium.

It found that an hour of cEEG monitoring was enough, when combined with the algorithm, to predict which acutely ill patients were likely to have seizures. Patients at higher risk could then be treated to prevent the brain injury that seizures can cause.

“Our ultimate goal is to help patients recover from their illness faster and improve their long-term neurological function,’’ Struck says.

Last year, Struck shared the INFORMS Analytics Society Innovative Applications in Analytics Award for his research in how to better predict seizures in patients with critical illness.