B(I)RD Watching: A Way to Stratify Seizure Risk?

Commentary Seizures are deceptively difficult to define. Clinical seizures with an EEG correlate are rarely problematic, but seizures can be purely clinical, with no scalp EEG correlate, or conversely, purely electrographic, with no apparent clinical correlate. The latter is a common occurrence in neurologically critical patients. There is recognition that seizure definitions set somewhat arbitrary cut points in a continuum that spans interictal and ictal findings. In reality, the determination of whether a given pattern is ictal is more a probabilistic one than a binary choice of ictal or interictal (1). Yet there are still sound reasons for establishing definitions of electrographic seizures. A common language is needed for clinical and research purposes. Definition of an EEG pattern as ictal often, but not always, signals changes in or intensification of treatment. Most attempts to define EEG criteria for seizures incorporate several features of the EEG (4). There is usually a requirement that the activity is rhythmic or periodic. Frequency is important, with frequencies of 3–4 Hz or greater usually representing ictal patterns, with the understanding that lower frequency discharges are also potentially ictal. Evolution, or change over time, is characteristic of most seizures, and evolution in frequency, spatial characteristics, or morphology is a part of most definitions. Duration of the EEG pattern is often included but admittedly an arbitrary criterion. While it may not make biological sense to set a lower bound for the length of a seizure, there are practical issues in distinguishing brief rhythmic runs from more fully elaborated seizures, and a lower limit of 10 seconds duration is often used. It is the sub-10-second realm that Yoo and coworkers explore in their article. In doing so, they coin a new term: B(I)RDs. The acronym stands for " brief, potentially ictal rhythmic discharges, " with the parenthetical (I) indicating that the discharges are potentially, but not definitely ictal. They defined B(I)RDs as very brief (<10 second) lateralized runs of rhythmic activity with a frequency greater than 4 Hz. For these brief discharges, features of evolution were not required. Similar patterns have been previously described in neonates. The authors set out with several goals: to report the prevalence, the significance, and the prognostic implications of these brief rhythmic discharges in adults. To do so, Yoo and colleagues queried a large database of continuous EEG recordings at a single institution. The median age of subjects was 65 (range 13-88). Of …