Bilaterally suppressed EEG amplitudes predict death and poor functional outcomes in critically ill children

Abstract

Background and objectivesContinuous full-channel EEG is the gold standard for electrocortical activity assessment in critically ill children, but its implementation faces challenges, leading to a growing use of amplitude-integrated EEG (aEEG). While suppressed aEEG amplitudes have been linked to adverse outcomes in preterm infants and adults after cardiac arrest, evidence for critically ill children remains limited. This retrospective study aimed to evaluate the association between suppressed aEEG amplitudes in critically ill children and death or poor functional neurological outcomes.Methods235 EEGs derived from individual patients < 18 years in the pediatric intensive care unit (PICU) at the University Hospital Essen (Germany) between 04/2014 and 07/2021 were retrospectively converted into aEEGs and amplitudes analyzed with respect to previously defined age-specific percentiles. Adjusted odds ratios for death and poor functional outcome at hospital discharge in patients with bilateral upper or lower amplitude suppression below the 10thpercentile were calculated accounting for neurological injuries, acute disease severity, sedation levels, and functional neurological status before acute critical illness.ResultsThe median time from neurological insult to EEG recording was 2 days. PICU admission occurred due to neurological reasons in 43 % and patients had high overall disease severity. Thirty-three (14 %) patients died and 68 (29 %) had poor outcomes. Amplitude depression below the 10thpercentile was frequent (upper amplitude: 27 %, lower amplitude: 34 %) with suppression of only one amplitude less frequent than bilateral suppression. Multivariable regression analyses yielded odds between 6.63 and 15.22 for death, neurological death, and poor neurological outcomes if both upper or both lower amplitudes were suppressed. Model discrimination was excellent with areas under the curve above 0.92 for all models.DiscussionThis study found a high prevalence of suppressed aEEG amplitudes in critically ill children early after PICU admission, with suppression being highly associated with death and poor functional outcomes at hospital discharge. These findings emphasize the potential of early identification of high-risk PICU patients through aEEG monitoring if conventional EEG is unavailable, potentially guiding neuroprotective therapies and early neurorehabilitation.