Early clinical predictors of a severely abnormal amplitude-integrated electroencephalogram at 48 hours in cooled neonates

Aim There is a need to identify infants with hypoxic ischaemic encephalopathy who have a poor outcome despite therapeutic hypothermia. A severely abnormal amplitude-integrated electroencephalogram at 48 h predicts death or disability. Our aim was to determine whether clinical assessment at age 3–5 h predicts a severely abnormal amplitude-integrated electroencephalogram at 48 h or death in cooled infants. Methods Forty-one cooled infants, ≥36 weeks' gestation, with moderate-to-severe hypoxic ischaemic encephalopathy, were prospectively enrolled. Infants who were moribund, had congenital conditions associated with encephalopathy or had severe cardio-respiratory instability were excluded. The predictive abilities of the Thompson encephalopathy score and individual signs at age 3–5 h were assessed. Results All infants with a Thompson score ≥16 at 3–5 h had a severely abnormal amplitude-integrated electroencephalogram at 6 h and an abnormal short-term outcome. At 48 h, 75% had a severely abnormal aEEG or died vs. 18% with a score <16 (p = 0.004). Multivariate analysis did not find a significant independent association with any of the individual signs. Conclusion The Thompson score could be useful to identify infants who will have a poor outcome despite cooling. A score ≥16 should be validated as a prespecified variable in prospective studies.