Our objective is to integrate the bilirubin risk prediction model into the automated bilirubin assessment in the laboratory results and provide follow-up guidance on timing of repeat bilirubin testing. We hypothesize that integration of this model will lead to better risk- stratification and a decline in bilirubin testing without an increase in readmissions for phototherapy or severe hyperbilirubinemia. Our specific aims are to compare the 1) mean number of bilirubin results per infant 2) phototherapy readmissions and 3) incidence of severe hyperbilirubinemia among newborns managed using the bilirubin prediction model and automated follow-up recommendations compared to newborns managed at the discretion of individual physicians.
Predictive Analytics for Newborn Bilirubin Management
Investigator: Kuzniewicz, Michael
Funder: Northern California Community Benefit Programs