Construction and verification of neonatal ABO-blood group system hemolytic disease of the newborn risk prediction model
Objective To construct a model that can predict the risk of diagnosing ABO-blood group sys-tem hemolytic disease of the newborn(ABO-HDN)and to verify its effectiveness.Methods A total of 446 children with neonatal hyperbilirubinemia who met the inclusion criteria and were first diagnosed in this hos-pital from January 2022 to March 2023 were selected as the modeling group,and were divided into the develo-ping group(200 cases)and the non-developing group(246 cases)according to whether ABO-HDN was diag-nosed.Totally 17 potential influencing factors were included for univariate analysis and multi-factor analysis,and independent risk factors were included in R software to establish a Nomogram model to predict the risk of ABO-HDN.Another 105 cases of neonatal hyperbilirubinemia in the hospital from April to September 2023 were selected as the verification group.Results In the modeling group,Logistic regression analysis showed that maternal pregnancy number,prenatal serum titer,hemoglobin level,white blood cell count,creatine ki-nase level and neonatal Apgar 1 min score were all independent risk factors for ABO-HDN(P＜0.05).Multi-variate Logistic regression analysis showed that the area under receiver operating characteristic(ROC)curve of the modeling group was 0.819(95％CI:0.779-0.859),sensitivity was 0.655,specificity was 0.878.In the verification group,the area under ROC curve was 0.867(95％CI:0.800-0.933),the sensitivity was 0.803,and the specificity was 0.773.Conclusion The established predictive model scoring system can effec-tively predict the risk of ABO-HDN.
ABO blood typehemolytic disease of the newbronrisk factorsprediction modelexternal verification
万方数据
