A Prediction Nomogram for Moderate‐to‐Severe Bronchopulmonary Dysplasia in Chinese Very Preterm Infants With Respiratory Distress: A Multi‑Center Prospective Study

This is the authors' abstract. We don't add key points for this paper.

Moderate-to-severe bronchopulmonary dysplasia (msBPD) is a prevalent and severe condition in very preterm infants, contributing to long-term respiratory issues. Developing a predictive model specific to the Chinese population could support early intervention. This research focused on determining risk factors and developing a predictive model for msBPD in very preterm infants with respiratory distress, following the 2018 National Institute of Child Health and Human Development criteria. A prospective multicenter study was conducted across 26 tertiary neonatal intensive care units (NICUs) in China from March 2020 to March 2022. Infants born at < 32 weeks' gestation, admitted within 72 h of birth with a respiratory score ≥ 5, were enrolled. Infants were assigned at random to either a training cohort (800 infants from 18 NICUs) or an external validation cohort (397 infants from 8 NICUs). Predictive factors were identified using multivariate logistic regression and Lasso regression, leading to the development of a prediction nomogram. Key independent predictors included birth weight, surfactant use, early pulmonary hypertension, duration of invasive mechanical ventilation, and necrotizing enterocolitis (Stages II-III). The model demonstrated high accuracy in the training cohort (AUC = 0.844) with good calibration (Hosmer-Lemeshow test, p = 0.247). Independent validation showed consistent discrimination (AUC = 0.849) and calibration (Hosmer-Lemeshow test, p = 0.333), while decision curve analysis confirmed clinical benefit. The developed nomogram-based model offers reliable early msBPD risk assessment in very preterm infants, supporting timely clinical interventions within Chinese NICUs.