Qualitative aEEG Analysis in Hypoxic-Ischaemic Encephalopathy in the First Six Hours After Birth

Chiara Lamond, Denise Chan, Timothy Schindler, John Smyth

Aim:

Hypoxic-ischaemic encephalopathy (HIE) accounts for 23% of global neonatal deaths, however, early classification of its severity remains difficult within the critical six-hour window required for effective treatment with therapeutic hypothermia (TH). The study aims to determine the use of qualitative analysis of amplitude-integrated electroencephalography (aEEG) in the first six hours after birth. In particular, if it can predict cerebral magnetic resonance imaging (MRI) findings in the first week and neurodevelopmental outcome at 12 months of age in infants with HIE to aid treatment decisions and prognostication.

Method:

This is a single-centre, retrospective observational study with visual assessment of one-hour epochs of aEEG in 19 infants with HIE who underwent TH. Two assessors independently analysed the aEEG with high interrater reliability. The primary outcome was assessing neurodevelopment at 12 months using Bayley’s Scales of Infant Development III. The secondary outcome was the severity of injury on brain MRI in the first week (Weeke score). The predictive ability of aEEG analysis was calculated.

Results:

Background pattern grading on aEEG significantly predicted an abnormal MRI and adverse neurodevelopmental outcome (MRI: p<0.001, neurodevelopment: p<0.005). Voltage pattern grading on aEEG was associated with an abnormal MRI (p<0.011) but did not reach significance in regard to neurodevelopmental outcome (p<0.053). A very severely abnormal background pattern was strongly associated with both an abnormal MRI and abnormal neurodevelopmental outcome (p<0.005). The modified-Sarnat score was not significant in predicting an adverse outcome (p<0.086).

Conclusions:

Early aEEG voltages and background patterns proved a better predictor of short and long-term outcome than the modified-Sarnat classification and may assist in selection of infants with HIE that are likely to benefit from TH within the critical window. A larger sample size is required to corroborate these findings.