Duration of nonconvulsive seizures in critically ill patients

Abstract

IntroductionNon-convulsive seizures (NCSz) and non-convulsive status epilepticus (NCSE) are frequent in critically ill patients. Specific temporal thresholds to define both are lacking and may be needed to guide appropriate treatment. MethodRetrospective review of 995 NCSz captured during continuous EEG monitoring of 111 consecutive critically ill patients. Seizures were classified according to their type and underlying etiology (acute or progressive brain injury, seizure-related disorders and acute medical illness). Median and interquartile ranges [IQR] were calculated. Suggested temporal threshold for NCSE was defined as the 95 percentile of seizure duration. ResultsMost (69%) patients had an underlying acute or progressive brain injury. The 95 percentile of seizure duration was 518 s, overall, with variation according to underlying etiology (median 86 [47–137] s for brain injury, 73 [45–115] s for seizure-related disorders, and 92 [58–223] s for acute medical illness, respectively; p = 0.0025; 95 percentile 424, 304, and 1725 s, respectively). Forty-one (37%) patients were comatose and had significantly longer seizures than non-comatose patients (median 99 [49–167] vs. 73 [46–123] s; p < 0.001; 95 percentile: 600 vs 444 s). ConclusionTo define NCSE, a temporal threshold of 10 min in critically ill patients with a primary neurological diagnosis can be applied, while a temporal threshold of 30 min might be suitable for patients with an underlying acute medical illness.