Risk analysis of adverse events in neonatal hyperbilirubinemia and establishment of a predictive model
A line graph model was developed to predict the risk of adverse events in neonatal hyperbilirubinemia(NHB),which is difficult to predict.Firstly,we collected relevant factors that might influence the occurrence of NHB adverse events,grouped children according to the presence or absence of brain injury and myocardial damage,performed multifactorial analysis,established a line graph model and performed internal validation.The results showed that a total of 34(17.99％)of 183 children with NHB had adverse events.Week of gestation at birth,feeding method,number of deliveries,TBIL,IBIL,and blood lactate were independent influencing factors for adverse events in children with NHB.Repeated sampling using Bootstrap method,calibration curve result display and ROC results suggest that the column line graph model predicts the risk of adverse events in children with NHB with high accuracy and discriminat.
万方数据
维普
