Prediction of Neurodevelopmental Outcomes at 18 to 22 Months Using the Numerical Sarnat Score Compared with Modified Sarnat Staging in Infants with Moderate to Severe Hypoxic-Ischemic Encephalopathy

Abstract

Objective To examine the association of a numerical Sarnat score (NSS) and modified Sarnat staging among newborn infants with moderate/severe hypoxic-ischemic encephalopathy and their neurodevelopmental outcomes at 18-22 months. Study design This secondary analysis included participants with gestational age >= 36 weeks and moderate/severe hypoxic-ischemic encephalopathy (HIE) from the Induced Hypothermia and Optimizing Cooling trials. Early (<6 hours age) neurologic examinations were performed by trained examiners to categorize HIE severity by modified Sarnat staging. The NSS was calculated by summing abnormal scores (2 for moderate, 3 for severe) in the 6 examination categories. The primary outcome was death or moderate/severe disability. Statistical analysis included logistic regression, adjusting for center, trial, and cooling group, and linear regression for continuous scales. Results The cohort (n = 528) included infants with 71% moderate and 29% severe HIE (37% Induced Hypothermia and 63% Optimizing Cooling participants). Median (IQR) NSS of infants with moderate and severe HIE were 11 (9-13) and 16 (16-17), respectively. There were significant associations among NSS, NSS tertiles, and modified Sarnat staging and death or disability, but there were no differences found in their area-under-the-curve estimates. Similar to modified Sarnat staging, NSS showed significant associations with Bayley-III cognitive, language, and motor scores and Bayley-II mental and psychomotor developmental indices. Conclusion Among infants with moderate or severe HIE, the NSS did not improve the predictive accuracy for death or disability at 18-22 months of age, compared with modified Sarnat staging performed in the initial 6 hours after birth by trained examiners.