Risk factors for early diagnosis and prediction model development of of neonatal ABO-HDFN
10.13303/j.cjbt.issn.1004-549x.2025.07.004
- VernacularTitle:胎儿新生儿ABO溶血病早期诊断危险因素分析及风险预测模型构建
- Author:
Wenhua ZHANG
1
;
Dan LIU
1
;
Wenting ZHANG
2
;
Jing LING
1
Author Information
1. Department of Transfusion Medicine, Children's Hospital of Soochow University, Suzhou 215000, China
2. Department of Cardiothoracic Surgery, Children's Hospital of Soochow University, Suzhou 215000, China
- Publication Type:Journal Article
- Keywords:
ABO-HDFN;
risk factors;
predictive model;
performance validation
- From:
Chinese Journal of Blood Transfusion
2025;38(7):886-895
- CountryChina
- Language:Chinese
-
Abstract:
Objective: To investigate the risk factors affecting the early diagnosis of ABO-hemolytic disease of the fetus and newborn (ABO-HDFN) in neonates with maternal-fetal blood group incompatibility, and to develop a risk prediction model and validate its predictive performance, so as to provide a reference for the early diagnosis of neonates with ABO-HDFN in primary hospitals. Methods: A total of 1 229 neonates with maternal-fetal blood group incompatibility suspected of ABO-HDFN, admitted to our hospital between between June 2021 and September 2024, were enrolled. The sample size was calculated by using the events per variable (EPV) method. The cohort was divided into a modeling group (n=860) and a validation group (n=369), and the results and clinical information of laboratory examination indicators were collected. Univariate and multivariate logistic regression analysis were used to explore the risk factors affecting the early diagnosis of ABO-HDFN in neonates with maternal-fetal blood group incompatibility. The risk prediction model was developed and internally validated by the Bootstrap method. The goodness-of-fit of the model was evaluated by the Hosmer-Lemeshow (H-L) test, and the receiver operating characteristic (ROC) curve was used to analyze the predictive performance of the model. The prediction model was validated by using the validation group data, and the predictive performance of the model was evaluated. Results: Among the 860 neonates with maternal-fetal incompatibility in the modeling group, 346 (346/860, 40.23%) were diagnosed with ABO-HDFN. Univariate and multivariate logistic regression analyses identified the following as significant risk factors for early diagnosis: the number of postnatal days at specimen collection, maternal type O blood group, parity >1, time of onset for pathologic jaundice, maternal-fetal blood group incompatibility due to A antigen, the level of total bilirubin, and the immature reticulocyte fraction (IRF). A risk prediction model was established, and the calibration degree of the model was validated by the Bootstrap internal validation method, Brier=0.143. The results of H-L test showed that χ
=3.464, P=0.902. The area under the ROC curve (AUC) was 0.885. The maximum value of the Youden index was 0.611, the sensitivity was 0.832, and the specificity was 0.778. The results of the validation group showed that the area under the ROC curve was 0.863, with a sensitivity of 0.875 and specificity of 0.735. Conclusion: The risk prediction model developed based on these risk factors has good predictive performance for ABO-HDFN, facilitating early diagnosis of suspected ABO-HDFN cases by clinicians in primary hospitals.
