Association of early biochemical markers with in-hospital mortality in infants with neonatal encephalopathy.

Abstract

Our aim was to describe the possible association between early, routinely measured serum laboratory parameters and in-hospital mortality in neonates with neonatal encephalopathy (NE), treated with hypothermia. This retrospective exploratory cohort study analyzed the results of 19 routinely measured laboratory parameters taken at 6 h of postnatal life from neonates with NE. The primary outcome was defined as in-hospital mortality. A total of 441 neonates were analyzed, the rate of in-hospital mortality was 12.5%. Multivariable logistic regression analysis (adjusted for 5 min Apgar, initial base excess and Thompson score) showed that for every 1 mmol/L increase in serum phosphate level, the odds of death increased by 8.43-fold (95% CI 3.33-21.39; p = 0.001). ROC analysis demonstrated that a serum phosphate level above 2.58 mmol/L (a threshold remaining within the normal range) provided strong discriminatory potential for in-hospital mortality within this cohort (AUC 0.90; sensitivity 80%, specificity 89%). Among survivors, neurodevelopmental impairment (using the Bayley-II test) was not associated with the initial serum phosphate level. Our results suggest that early serum phosphate may be a potential biomarker in the future and may additionally be used to determine the severity and short-term prognosis of neonates with NE. Increased serum phosphate level at six hours of postnatal age was associated with higher odds of in-hospital mortality in infants with neonatal encephalopathy. Unlike for mortality prediction, phosphate was not a significant predictor of neurodevelopmental outcomes. Early serum phosphate may be a potential biomarker in the future and may additionally be used to determine the severity and short-term prognosis of neonates with neonatal encephalopathy.