Development and Validation of a Seizure Prediction Model in Neonates After Cardiac Surgery

Electroencephalographic seizures (ESs) after neonatal cardiac surgery are often subclinical and have been associated with poor outcomes. An accurate ES prediction model could allow targeted continuous electroencephalographic monitoring (CEEG) for high-risk neonates. ES prediction models were developed and validated in a multicenter prospective cohort where all postoperative neonates who underwent cardiopulmonary bypass (CPB) also underwent CEEG. ESs occurred in 7.4% of neonates (78 of 1053). Model predictors included gestational age, head circumference, single-ventricle defect, deep hypothermic circulatory arrest duration, cardiac arrest, nitric oxide, extracorporeal membrane oxygenation, and delayed sternal closure. The model performed well in the derivation cohort (c-statistic, 0.77; Hosmer-Lemeshow, P = .56), with a net benefit (NB) over monitoring all and none over a threshold probability of 2% in decision curve analysis (DCA). The model had good calibration in the validation cohort (Hosmer-Lemeshow, P = .60); however, discrimination was poor (c-statistic, 0.61), and in DCA there was no NB of the prediction model between the threshold probabilities of 8% and 18%. By using a cut point that emphasized negative predictive value in the derivation cohort, 32% (236 of 737) of neonates would not undergo CEEG, including 3.5% (2 of 58) of neonates with ESs (negative predictive value, 99%; sensitivity, 97%). In this large prospective cohort, a prediction model of ESs in neonates after CPB had good performance in the derivation cohort, with an NB in DCA. However, performance in the validation cohort was weak, with poor discrimination, poor calibration, and no NB in DCA. These findings support CEEG of all neonates after CPB.