NCT01109277

Prediction of Neonatal Hyperbilirubinemia

An Evidence-based Strategy for Assessing the Risk of Significant Neonatal Hyperbilirubinemia

• Status: Completed
• Phase: —
• Study type: Observational
• Enrollment: 3,500 (estimated)
• Sponsor: University of Patras · Academic / Other
• Sex: All
• Age: 1 Hour – 15 Days
• Healthy volunteers: Not accepted

Summary

Objective: To develop an evidence-based strategy for assessing the risk of significant hyperbilirubinemia in healthy term and near-term (late-preterm) neonates. Hypothesis: A stepwise strategy which combines clinical parameters and serial non-invasive transcutaneous bilirubin (TcB) values could reliably predict significant neonatal hyperbilirubinemia. Methods: Data from neonates \>34 weeks' gestation included in the registry for neonatal hyperbilirubinemia of the well-baby nursery of the University Hospital of Patras, from January 2008 to December 2010 will be reviewed. The registry includes prospectively collected data such as sex, gestational age, gestation and perinatal information, mother's and infant's ABO group and Rh, G6PD deficiency, Coombs test, type of delivery and complications, birthweight, postnatal medications and interventions, type and volume of feeding (daily), extension of jaundice, TcB measurements at intervals of 12+/-4 hours until discharge, total serum bilirubin values (if obtained), TcB or TSB measurements at follow-up, weight at discharge, need of phototherapy (inpatient or after discharge). TcB and TSB values are plotted on a hour-specific chart. A novel predictive nomogram based on TcB measurements (Varvarigou et al. Pediatrics 2009;124:1052-9) will be used to classify TcB values as high, intermediate, and low risk. Significant hyperbilirubinemia will be defined as a TSB value above the phototherapy threshold level according to the AAP 2004 guidelines Statistics: Independent and joint effects of various clinical factors on the development of significant hyperbilirubinemia will be evaluated by logistic regression analysis Cluster analysis and Chi-squared Automatic Interaction Detection (CHAID) tree method will be used to develop the strategy. At each step, CHAID chooses the independent (predictor) variable that has the strongest interaction with the dependent variable. Categories of each predictor are merged if they are not significantly different with respect to the dependent variable.

Conditions

• Hyperbilirubinemia, Neonatal

Timeline

• Start date: 2010-04-01
• Primary completion: 2011-01-01
• Completion: 2011-01-01
• First posted: 2010-04-23
• Last updated: 2011-01-19

Locations

1 site across 1 country: Greece