No abstract

No abstract

BackgroundBilirubin-related neurotoxicity is an important clinical issue in very low birthweight (VLBW) infants, and the existing literature is inconsistent.ObjectiveTo analyze the relationship between maximal serum unconjugated bilirubin levels (SBL) and neurodevelopmental outcome at 2-year corrected age in VLBW infants.MethodsPhototherapy was initiated in all infants born before 33 weeks of gestation, according to Maisels' recommendations. Neurodevelopmental assessment at 2-year corrected age was performed in all infants that survived. SBLs collected during the first week of life were used to define three tertiles of max-SBL. The first tertile corresponded to infants with the lowest max-SBL.Results and ConclusionsA total of 724 infants were included in the study, and among them, 631 (87%) were evaluated at two years old. The infants of the first tertile were younger and smaller than the infants of the other two tertiles, in accordance with Maisels' recommendations for very small infants. No difference in the risk of impaired functional outcome among the three groups was observed. However, among infants weighing less than 1001 g, those in the third tertile had a poorer neurodevelopmental prognosis as compared to those in the second tertile (adjusted odds ratio = 6.8, 95% CI: 1.2–36.7, p = 0.03). Considering the results obtained, we propose 196 µmol/L (11.5 mg/dL) when birthweight varies between 1001 and 1500 g, and 170 µmol/L (9.9 mg/dL) when birthweight is less than 1001 g, as recommended max-SBLs (defined as maximal levels of 95th percentile curves of SBLs in infants with an optimal outcome). When Maisels' recommendations were applied, max SBLs were higher in 8% of infants weighing 1001–1500 g and in 15% of infants weighing less than 1001 g. Our data seems to validate Maisels' recommendations in the overall population of infants born before 33 weeks of gestation, but not in infants weighing less than 1001 g.

We evaluated the semi-quantitative procalcitonin level for diagnosing late-onset infections in 176 neonates. Using a cut-off level of 0.5 ng/ml, the sensitivity was 84.4%+/-0.19, specificity was 93.9%+/-0.04, positive predictive value was 82.6%+/-0.1, and negative predictive value was 94.6%+/-0.04. Procalcitonin could be a useful marker of late-onset infection in neonates.

No abstract