Prognostic model for predicting survival in very preterm infants: an external validation study.

Area of Science:

• Neonatal Medicine
• Clinical Epidemiology
• Biostatistics

Background:

• A high-quality prognostic model for predicting survival in very preterm infants was previously developed.
• This model requires external validation to assess its generalizability and reliability.

Purpose of the Study:

• To perform temporal and geographical validation of a prognostic model for predicting survival in very preterm infants.
• To assess the model's performance in an independent, population-based cohort.

Main Methods:

• External validation study using a population-based cohort of infants born between 23 and 32 weeks of gestation.
• Model performance evaluated using calibration (calibration-in-the-large, calibration slope) and discrimination (c statistic).
• Model was applied to multiple validation sets, including singletons and multiples, with potential updates.

Main Results:

• The model demonstrated good external validity with strong discrimination (c statistic 0.82) and calibration.
• Survival rate in the cohort was 92.9% (5659 out of 6092 infants).
• The model performed well in singleton populations and required a minor update for multiples.

Conclusions:

• A high-quality prognostic model for very preterm infant survival is externally valid in an independent, nationwide cohort.
• The model's accurate performance supports its potential implementation in neonatal intensive care unit clinical practice.
• Further impact assessment is recommended before widespread clinical adoption.