Prevalence and Predictors of Seizure Clusters in Pediatric Patients With Epilepsy: The Harvard-Yale Pediatric Seizure Cluster Study.

ABSTRACT

BACKGROUND: Determine the prevalence of seizure clusters (two or more seizures in six hours), use of rescue medications, and adverse outcomes associated with seizure clusters in pediatric patients with a range of epilepsy severities, and identify risk factors predictive of seizure clusters. METHODS: Prospective observational two-center study, including phone call and seizure diary follow-up for 12 months in patients with epilepsy aged one month to 18 years. We classified patients into three risk groups based on seizures within the prior year high, seizure cluster (two or more seizures within one day); intermediate, at least one seizure but no days with two or more seizures; low, no seizures. RESULTS: One-third (32.3%; high risk, 72.4%; intermediate risk, 30.4%; low risk, 3.1%) of 297 patients had a seizure cluster during the study, including half (46.2%) of the patients with active seizures at baseline (intermediate- and high-risk groups combined). Emergency room visits or injuries were no more likely due to a seizure cluster than an isolated seizure. Rescue medications were utilized in 15.8% of patients in the high-risk group and 19.2% in the intermediate-risk group. History of status epilepticus (adjusted odds ratio [aOR], 2.13; confidence interval [CI], 1.09 to 4.16]), seizure frequency greater than four per month (aOR, 4.27; CI, 1.92 to 9.50), and high-risk group status (aOR, 6.42; CI, 2.97 to 13.87) were associated with greater odds of seizure cluster. CONCLUSIONS: Seizure clusters are common in pediatric patients with epilepsy. High seizure frequency was the strongest predictor of clusters. Rescue medications were underutilized. Future studies should evaluate the applicability and effectiveness of these medications for optimization of pediatric seizure cluster treatment and reduction of seizure-related emergency department visits, injuries, and mortality.