Related Experiment Video
Updated: Jan 8, 2026
Establishing a Competing Risk Regression Nomogram Model for Survival Data
Published on: October 23, 2020
Development and internal validation of a clinical nomogram for predicting bronchopulmonary dysplasia in preterm
Yan-Sha Pan1, Lan Xiao1, Wen-Bin Dong1
1Department of Pediatrics, Sichuan Clinical Research Center for Birth Defects, The Affiliated Hospital, Southwest Medical University, Luzhou, Sichuan 646000, P.R. China.
Insights
This study developed a clinical prediction model to assess the risk of bronchopulmonary dysplasia (BPD) in preterm infants. The model accurately identifies key predictors, enabling early intervention for this common neonatal condition.
Area of Science:
- Neonatal Medicine
- Pediatric Pulmonology
- Clinical Prediction Modeling
Background:
- Bronchopulmonary dysplasia (BPD) is a significant cause of illness in premature infants.
- Early identification of infants at risk for BPD is crucial for timely interventions.
Purpose of the Study:
- To develop and internally validate a clinical prediction model for BPD in preterm infants.
- To identify independent predictors of BPD for risk stratification.
Main Methods:
- Retrospective analysis of 120 preterm infants (<32 weeks gestation).
- Classification into BPD and non-BPD groups using 2018 NICHD criteria.
- Development of a nomogram using logistic regression, validated with bootstrapping.
Main Results:
- Gestational age, birth weight, sepsis, PDA, and IVH were identified as independent BPD predictors.
- The model showed strong discrimination (AUC=0.918) and good calibration.
- The nomogram provided individualized BPD risk estimation with confirmed robustness.
Conclusions:
- The developed nomogram is a robust tool for early BPD risk assessment in preterm infants.
- The model demonstrates significant clinical applicability for guiding interventions.
- Further multicenter validation is recommended to enhance generalizability.
Abstract:
Bronchopulmonary dysplasia (BPD) is a major morbidity in preterm infants, necessitating early risk assessment to guide interventions. The present study aimed to develop and internally validate a clinical prediction model for BPD. A total 120 preterm infants (<32 gestation weeks) admitted to a neonatal intensive care unit from January 2020 to December 2022 were retrospectively analyzed. Infants were retrospectively classified into BPD (n=34) and non-BPD (n=86) groups based on the 2018 National Institute of Child Health and Human Development criteria. Clinical variables, including maternal, neonatal, respiratory and comorbid factors, were assessed. Univariate and multivariate logistic regression identified independent predictors, which were used to construct a nomogram. Model performance was evaluated using the area under the curve (AUC) of a receiver operating characteristic curve, a calibration curve and Hosmer-Lemeshow test. Internal validation was performed via bootstrapping. The results demonstrated that gestational age, birth weight, sepsis, patent ductus arteriosus and intraventricular hemorrhage were independent predictors of BPD. The model demonstrated good discrimination (AUC=0.918; 95% confidence interval, 0.866-0.971) and good calibration. The nomogram enabled individualized risk estimation, and internal validation confirmed model robustness. In conclusion, the proposed nomogram demonstrated strong discriminative power and clinical applicability for early BPD risk assessment. Future multicenter validation will help extend its generalizability across diverse neonatal populations.
