Video electroencephalograms are more likely to capture an event of interest if there have been quite frequent past events, if the last event was within 24 hours of admission, and if the patient has an intellectual disability, according to findings from a retrospective study of 213 children who underwent prolonged inpatient monitoring during a recent 3-year period at one institution.
The procedure probably wouldn’t be as fruitful for developmentally normal children or for those who have less frequent spells with more time between them, Kirk D. Wyatt and his coauthors at the Mayo Clinic, Rochester, Minn., reported (Pediatr. Neurol. 2014 Jan. 27 [doi:10.1016/j.pediatrneurol.2014.01.038]).
The findings of their study can be used to counsel families about the likelihood of having a successful admission – and even as a basis to suggest that they try to capture an event on their own, said the investigators, who were led by senior author Dr. Elaine C. Wirrell, a professor of neurology and director of the pediatric epilepsy fellowship program at Mayo.
Dr. Elaine Wirrell
"The utility of [video EEG] monitoring for events that do not occur at least on a weekly basis is limited, even in the context of identifiable provocative factors," the colleagues wrote. "With the ubiquity of cellular phone video cameras, asking parents to make a recording of a rare event when it occurs in vivo so that a child neurologist may review it in conjunction with the remainder of the clinical history and a routine EEG may be higher yield than video EEG monitoring, at least as an initial step."
The inpatient stay of the children, aged 2-13 years, was intended to capture a physical event related to EEG changes. Overall, the median recording duration was 25 hours, but that ranged from 22 to 48 hours.
The procedure captured at least one event in 66% of the children at a median monitoring duration of 4.5 hours. The median time to capturing an EEG-related event was directly related to how often they occurred before admission. For those who had daily events, the median time to capture was about 4 hours. That jumped to 24 hours for children who had events at least three times a week, and to almost 23 hours for those who had them once or twice a week. The median time to capture was about 8 hours for children whose events occurred less than once a week.
Events that occurred most frequently (at least three times a week) were recorded during admission 72% of the time. That dropped to 41% of the time for events that occurred at frequencies ranging from less than three times a week to monthly, and 26% of the time for those that occurred less than monthly.
Getting a child in for monitoring soon after an event also affected the success of the procedure. The median time to capture was almost 4 hours if the last event happened less than 24 hours before admission, but it jumped to 22.4 hours if the last event had happened 24-72 hours prior and to 22.7 hours of it had occurred between 72 hours and 1 week earlier. But the median time to capture was nearly 15 hours if the last event had occurred more than a week before admission.
Events were captured 71% of the time when the last event had occurred less than 24 hours before admission. Success dropped to 52% for events that had occurred 24-72 hours before admission and to 32% when they had occurred more than 72 hours before admission.
The highest preadmission event frequency significantly increased the chance of capturing an event (odds ratio, 3.77), as did the shortest event latency (OR, 2.31).
Intellectual disability in the patient significantly increased the likelihood of capturing an EEG-related event (OR, 3.26). But the common practices of sleep deprivation and antiepileptic medication withdrawal didn’t increase the likelihood of capturing one, the authors noted. In fact, medication withdrawal or dose change actually decreased the chance of an event capture (OR, 0.46).
That finding "came as a surprise," the investigators noted, but it might have been spurious, because it was not an a priori outcome and because there were not many patients in these subgroups. However, they added, "It typically takes five half-lives to reach a new steady state of each medication adjustment. Therefore, depending upon the half-life duration of the medications, levels may not become subtherapeutic for several days after discontinuation or dose reduction."
A family history of epilepsy, a prior interictal discharge on routine EEG, and a prior diagnosis of possible epilepsy all also significantly increased the chance that a captured event would be EEG related.