An Algorithm for Predicting Neonatal Mortality in Threatened Very Preterm Birth

Area of Science:

• Perinatal Medicine
• Neonatal Health
• Predictive Modeling

Background:

• Neonatal mortality remains a significant concern in preterm infants.
• Antenatal information offers potential for early risk assessment.
• Predictive models can improve obstetric management and counseling.

Purpose of the Study:

• To develop a predictive model for neonatal mortality.
• Utilize information available during the antenatal period.
• Identify key antenatal factors associated with neonatal mortality.

Main Methods:

• A multiple logistic regression model was applied.
• A population-based cohort of very preterm infants (23+0 to 30+6 weeks' gestation) was analyzed.
• Infants under 23 weeks and those with major anomalies were excluded.

Main Results:

• Decreasing gestational age was a strong predictor of increased mortality.
• Factors like small for gestational age, oligohydramnios, maternal psychiatric disorders, antenatal antibiotics, and monochorionic twins increased mortality risk.
• Antenatal antihypertensives and corticosteroids (≥24 hours pre-delivery) were associated with reduced mortality.

Conclusions:

• Neonatal mortality prediction is influenced by maternal and fetal factors.
• An accessible algorithm was developed to estimate mortality risk.
• This tool can facilitate counseling and shared decision-making in obstetric care.