Comparison of numerical and standard sarnat grading using the NICHD and SIBEN methods

The NICHD and SIBEN assessments are adapted from the Sarnat grade, and used to determine severity of neonatal encephalopathy (NE). We compare NICHD and SIBEN methods, and their ability to define a minimum threshold associated with significant cerebral injury.

Between 2016 and 2019, 145 infants with NE (77-mild; 65-moderate; 3-severe) were included. NICHD and SIBEN grade and numerical scores were assigned. Kappa scores described agreement between methods, and ROC curves their ability to predict MR injury.

Good agreement existed between grading systems (K = 0.86). SIBEN defined more infants as moderate, and less as mild, than NICHD (p < 0.001). Both numerical scores were superior to standard grades in predicting MR injury.

Despite good agreement between methods, SIBEN defines more infants as moderate NE. Both numerical scores were superior to standard grade, and comparable to each other, in defining a minimum threshold for cerebral injury. Further assessment contrasting their predictive ability for long-term outcome is required.