A predictive model for infant mortality for 22-31 week gestation outborn and inborn livebirths

Background: Perinatal healthcare providers need access to contemporaneous outcome data to counsel parents and to inform decision-making regarding management of threatened preterm births. Our aim was to develop a model to predict mortality and survival to one year of age for outborn and inborn infants at each week of gestational age from 22 to 31 weeks.

Method: Using population-based data from 13,352 livebirths, free of lethal anomalies in Victoria sourced from the Consultative Council on Obstetric and Paediatric Mortality and Morbidity, we performed a multivariate logistic regression analysis to identify risk factors independently associated with mortality. Using the coefficients for the three variables independently associated with mortality (male sex, gestational age and inborn/outborn status) and the constant from the multiple logistic regression model, the probability of infant mortality for outborn males, outborn females, inborn males and inborn females at each week of gestational age from 22 to 31 weeks was calculated.

Results: When the predicted probability of death for outborn and inborn infants at each week of gestational was plotted with the actual mortality rates (1,777 infant deaths) the model accurately predicted mortality for outborn and inborn infants at each week of gestational age from 22–31 weeks.

Conclusions: We have developed a model to predict infant mortality for livebirths, free of lethal anomalies. This model predicts mortality for all livebirths, including delivery room deaths and is therefore more accurate than a model based on outcome data for infants admitted to a neonatal intensive care unit after birth.