To evaluate whether the Robson classification and obstetric risk profile can stratify the severity of care among neonates admitted to the NICU and identify factors associated with level 3 NICU care.
Approach:
Study Design: Retrospective, single-center study including 1,815 neonates admitted to the NICU from January 2023 to December 2025.
Data Analysis: Maternal, obstetric, and neonatal characteristics were analyzed across NICU levels using univariable and multivariable binary logistic regression.
Nomogram Development: A nomogram was constructed to estimate individualized risk, with model performance evaluated using discrimination and calibration methods.
Key Findings:
40.4% of NICU-admitted neonates required level 3 care.
Lower gestational age and birth weight were associated with higher NICU levels (p < 0.001).
Each additional week of gestation reduced the likelihood of level 3 care (OR = 0.82).
A history of miscarriage was associated with lower odds of level 3 care (OR = 0.15).
Non-low risk mothers according to Robson classification had higher odds of level 3 care (OR = 1.46).
The model demonstrated modest discrimination (C-index: 0.695).
Interpretation:
Gestational age, obstetric risk profile, and Robson classification contribute to stratifying the severity of care among NICU-admitted neonates.
Limitations:
The findings regarding history of miscarriage should be interpreted with caution.
The model is exploratory and requires external validation and refinement before broader clinical application.
Conclusion:
The proposed model serves as an exploratory risk-stratification framework rather than a clinically actionable prediction tool.
