A Retrospective Analysis of Current Clinical Utilization of Long-Term Video Electroencephalogram Monitoring at a...

Long-term video electroencephalogram monitoring (LTVM) is critical to the detection and management of seizures. Prior work has shown that most seizures are detected within 24 hours of monitoring. In practice, many patients are monitored for longer especially those who are at higher risk for seizures. This study examines whether higher risk clinical characteristics affect seizure detection rate.

A retrospective chart review was conducted on 847 LTVM studies of patients aged 0-18 years between August 1, 2021, and January 31, 2022, for those who were acutely hospitalized. Demographic, clinical, and electroencephalogram data were reviewed and analyzed. Descriptive statistics were used to analyze data, comparing seizure yield between clinical indications and age subgroups.

Median patient age was 1.5 years (interquartile range: 0.3-8.3 years). The most common indication for LTVM was seizure-like activity (51%, n = 429). The most common age group that utilized inpatient LTVM was infants (35%, n = 298). Seizures were detected in 22% (n = 189/847) of electroencephalograms. Median time to first detected seizure was 1.5 hours (interquartile range: 0.4-6.9 hours). Ninety-five percent of patients had seizures within 24 hours of LTVM initiation, regardless of age or indication. Comparison analysis revealed no statistical difference in overall seizure yield based on age or indication. High-risk subgroups of neonates, extracorporeal membrane oxygenation, and traumatic brain injury patients similarly did not have statistically different seizure yield on LTVM.

Almost all patients with seizures were detected within 24 hours regardless of age or clinical indication. These results indicate the need for standardization of LTVM resources.