What is the role of amplitude-integrated electroencephalography (aEEG) in the workup of hypoxic-ischemic encephalopathy (HIE)?

Updated: Jul 18, 2018
  • Author: Santina A Zanelli, MD; Chief Editor: Dharmendra J Nimavat, MD, FAAP  more...
  • Print
Answer

Several studies have shown that a single-channel aEEG performed within a few hours of birth can help evaluate the severity of brain injury in the infant with hypoxic-ischemic encephalopathy (HIE). [55, 56, 57] The abnormalities seen in infants with moderate-to-severe HIE include the following:

  • Discontinuous tracing characterized by a lower margin below 5 mV and an upper margin above 10 mV

  • Burst suppression pattern characterized by a background with minimum amplitude (0-2 mV) without variability and occasional high voltage bursts (>25 mV)

  • Continuous low voltage pattern characterized by a continuous low voltage background (< 5 mV)

  • Inactive pattern with no detectable cortical activity

  • Seizures, usually seen as an abrupt rise in both the lower and upper margin

In addition, aEEG findings have been used as criteria for inclusion in the CoolCap trial of selective head cooling. [27, 37, 58] However, some evidence argues against the use of aEEG as a tool to exclude infants with HIE from receiving hypothermia therapy.

Although normal aEEG findings may not necessarily mean that the brain is healthy, a severe or moderately severe aEEG abnormality may indicate brain injury and poor outcome. However, a rapid recovery (within 24 h) of abnormal aEEG findings is associated with favorable outcome in 60% of cases. Finally, in a meta-analysis of 8 studies, Spitzmiller et al concluded that aEEG can accurately predict poor outcome with a sensitivity of 91% (95% CI, 87-95) and a negative likelihood ratio of 0.09 (95% CI, 0.06-0.15). [59]

Note that considerable training is required for conducting and properly interpreting the aEEG findings. With more recent advancement in technology, aEEG can alert the caregivers regarding seizure activity based on the software recognition of a pattern.


Did this answer your question?
Additional feedback? (Optional)
Thank you for your feedback!