97 - EEG Monitoring in the Prediction of Outcome in Infants with HIE
Sunday, April 30, 2023
3:30 PM – 6:00 PM ET
Poster Number: 97 Publication Number: 97.336
Brian Walsh, University College Cork, Cork, Cork, Ireland; Carol Stephens, University College Cork, Cork, Cork, Ireland; Sean Mathieson, University College Cork, Rosscarbery, Cork, Ireland; Vicki Livingstone, UCC, Cork, Cork, Ireland; Deirdre Murray, University College Cork, Cork, Cork, Ireland; Geraldine B. Boylan, University College Cork, Cork, Cork, Ireland
Professor University College Cork Cork, Cork, Ireland
Background: Early prognostication is key for focused early interventions following HIE. EEG can be an important prognostication tool. Objective: To determine the optimal EEG timepoint in the first 72 hours after birth for the prediction of neurodevelopmental outcome in infants with HIE and to determine if EEG grading changed over this period and if that change differed by outcome group. Design/Methods: Infants were recruited as part of ANSeR, two large European multicentre studies conducted in eight sites. Inclusion criteria were infants ≥36 weeks with HIE, continuous EEG between 6-72 hours and outcome assessment at 2 years. EEG background was graded at 6, 12, 24, 36, 48, 60 and 72 hours. EEG was graded as normal (0), mild (1), moderate (2), severe (3) or inactive. For the purpose of analysis, grades ≤1 were classified as normal. The Area Under the Receiver Operating Characteristic curve (AUC) measured the predictive performance of EEG grades at each timepoint. Using linear mixed models, time was treated as a continuous variable and a straight line fitted to predict EEG grade for each group at any timepoint. The model investigated if EEG grading changed over the first 72 hours and whether changes differed by outcome group. Results: 185 infants were included; 60 mild HIE, 90 moderate HIE and 35 severe HIE. Fifty eight infants (31.4%) had an abnormal outcome. For both outcome groups, EEG evolution over the first 72 hours for each infant is described in figure 1. The most predictive timepoint for outcome was EEG grade at 24 hours; AUC(95%CI): 0.73(0.63 to 0.82). Infants with an abnormal EEG at this time were over three times more likely to have an abnormal outcome compared to those with a normal EEG at 24 hours; OR 3.48(1.73 to 7.01). EEG grades of 3 and 4 at any time (compared to a normal EEG grade across the 72 hours) were associated with a three and eight-fold increase respectively in the odds of an abnormal outcome. Changes in EEG grading over time did not differ between outcome groups (p =0.09 for group*time interaction). EEG grade was higher in the abnormal group compared to the normal group across all timepoints (difference in means(95% CI): 0.74 (0.50 to 0.97), p< 0.001). In both groups, background EEG improved over time with a decrease in EEG grade (difference in means(95%CI) for 72 hours vs 6 hours: -0.18 (-0.35 to -0.01), p=0.04). (Figure 2).
Conclusion(s): Background EEG grade at twenty-four hours of age is most predictive of outcome. EEG background grade improved over time but was always higher in the abnormal group. F500B3A8-8863-43CE-98B2-9D0FD13AE8B1.jpeg
