To explore early predictive value of scores of modified amplitude integrated electroencephalogram (aEEG) combined with serum level of neuron specific enolase (NSE) in newborns with neonatal bilirubin brain injury.

From March 2017 to February 2018, a total of 80 cases of newborns with hyperbilirubinemia who were hospitalized in neonatal ward of Northern Jiangsu People′s Hospital were selected as research subjects. According to results of head MRI examination of being present high signal changes in globus pallidus or not, they were divided into brain injury group (n=37) and non brain injury group (n=43). Meanwhile, there were 40 cases of healthy and full term newborns who born in the same hospital were selected as control group. For newborns in above 3 groups, their aEEG was monitored, and scores of aEEG were performed; serum concentration of total bilirubin (TBiL) and level of NSE, and brainstem auditory evoked potential (BAEP) were detected. The general clinical data, score of modified aEEG, serum concentration of TBiL and level of NSE of newborns in 3 groups were overall compared and further compared between each two groups by one-way ANOVA and least significant difference (LSD)-t test. Spearman rank correlation analysis was used to analyze correlation between score of modified aEEG, serum level of NSE, aEEG degree respectively, and BAEP degree. Receiver operating characteristic curve (ROC) was drawn to compare predictive value of score of modified aEEG, serum level of NSE and combination of these two indicators of bilirubin brain injury in neonates with hyperbilirubinemia. This study followed the ethical standards formulated by the Ethics Committee of our hospital, and was approved (Approval No. 2016KY-064). All guardians of included subjects were informed concents.

①There were no significant differences among 3 groups in gender composition, admission age and gestational age, weight, height, head and chest circumference at birth of newborns, and pregnancy and parity of their mothers (P>0.05). ② Scores of modified aEEG of newborns in brain injury group was (7.7±1.6) scores, which was lower than those of (10.4±1.5) scores and (10.9±1.3) scores in non brain injury group and control group, respectively; while serum concentration of TBiL and level of NSE were (349.7±74.0) μmol/L and (47.6±13.2) μg/L, respectively, which were significantly higher than those of (295.7±31.4) μmol/L and (32.3±7.7) μg/L in brain injury group, and those of (185.1±35.9) μmol/L and (28.6±7.3) μg/L in control group, and all the differences were statistically significant (all P<0.001). ③ For all subjects in this study, score of modified aEEG was negatively correlated with BAEP degree (r_s=-0.718, P<0.001), serum level of NSE and aEEG degree were both positively correlated with BAEP degree (r_s=0.685, P<0.001; r_s=0.647, P<0.001). ④ Area under curve of ROC (ROC-AUC) of score of modified aEEG, serum level of NSE and combination of these two indicators for prediction of bilirubin brain injury of neonates with hyperbilirubinemia were 0.875 (95%CI: 0.795-0.954, P<0.001), 0.853 (95%CI: 0.769-0.937, P<0.001) and 0.938 (95%CI: 0.861-0.980, P<0.001), respectively, the sensitivity of these three methods for prediction of brain injury of newborns were 86.5%, 62.2%, 83.8%, and the specificity were 74.4%, 93.0%, 90.7%, respectively.

The early predictive value of score of modified aEEG combined with serum level of NSE was higher than that of score of modified aEEG or serum level of NSE only.