Peer-reviewed article
Development and validation of a risk prediction model for preterm birth in women concurrent with Hepatitis B virus carrier status and gestational diabetes mellitus.
Zaidi Gui, Xiaoxiao Fan, Tianzi Zang, Hongyan Xu, Qi Wu et al. ·
Summary
Written by AI, not yet AICS's view · claude-haiku-4-5-20251001This study developed a prediction model to identify pregnant women at high risk of preterm birth who have both hepatitis B virus infection and gestational diabetes. Researchers analyzed data from 847 Chinese women with both conditions and identified 11 clinical and blood biomarker factors that predict preterm birth risk. The model showed good ability to distinguish between women who would and would not have preterm birth, suggesting it could help clinicians identify at-risk pregnancies for targeted intervention.
Key points
- 22% of women with concurrent hepatitis B virus carrier status and gestational diabetes mellitus experienced preterm birth
- Independent predictors included pre-pregnancy body mass index, blood pressure, liver enzymes, placenta previa, white blood cell count, protein, bile acid, thyroxine, membrane rupture mode, and prothrombin time
- The prediction model achieved an area under the curve of 0.854 in the training set and 0.754 in the validation set
- Decision curve analysis showed the nomogram provided favorable clinical utility for predicting preterm birth
- The model demonstrated good calibration, meaning predicted probabilities matched actual outcomes
Limitations
The study used retrospective data from a single hospital in one Chinese province, which may limit generalizability to other populations and healthcare settings. The external validation set showed lower discrimination (AUC 0.754) than the training set, suggesting the model's performance may vary in different clinical contexts.
Abstract
BACKGROUND: Preterm birth is a major adverse neonatal outcome that has been associated with maternal hepatitis B virus carrier status. Co-occurrence of hepatitis B virus carrier status carrier status and gestational diabetes mellitus is associated with an increased risk of preterm birth compared with hepatitis B virus carrier status alone. This study aimed to develop and externally validate a clinically applicable multivariable prediction model incorporating both clinical risk factors and maternal blood biomarkers to estimate preterm birth risk among women with concurrent hepatitis B virus carrier status and gestational diabetes mellitus. METHODS: A retrospective analysis was performed using clinical data and maternal blood biomarkers from 882 pregnant women with hepatitis B virus carrier status and gestational diabetes mellitus who delivered at Women's Hospital, in Zhejiang Province, China, from January 2015 to December 2024. Following data preprocessing, participants were randomly allocated to training and validation sets at a 7:3 ratio. Variable selection was performed using the least absolute shrinkage and selection operator (LASSO), and univariate and multivariable logistic regression analyses were subsequently conducted to assess the associations of clinical variables and maternal blood biomarkers with preterm birth in women with concurrent hepatitis B virus carrier status and gestational diabetes mellitus. A nomogram was developed to facilitate clinical application. Model discrimination and calibration were evaluated using receiver operating characteristic curve analysis and calibration plots, respectively, while clinical utility was assessed through decision curve analysis. RESULTS: A total of 847 women with hepatitis B virus carrier status and gestational diabetes mellitus were included in this study, of whom 187 (22%) experienced preterm birth. Multivariable logistic regression analysis identified pre-pregnancy body mass index, diastolic blood pressure, systolic blood pressure, alanine aminotransferase, placenta previa, white blood cell count, and maternal blood biomarkers-including total protein, total bile acid, total thyroxine, mode of membrane rupture, and prothrombin time-as independent predictors of preterm birth. These variables were incorporated into a nomogram, which demonstrated good calibration and discrimination. The area under the curve was 0.854 (95% CI: 0.823-0.881) in the training set and 0.754 (95% CI: 0.697-0.806) in the validation set. Decision curve analysis indicated that the nomogram provided favorable clinical utility for predicting preterm birth. CONCLUSION: The model demonstrated good discriminative ability and calibration, enabling accurate prediction of preterm birth risk in women with concurrent hepatitis B virus carrier status and gestational diabetes mellitus.