RESUMO
Preterm neonatal mortality contributes significantly to the high incidence of death among children under five years of age. Neonatal mortality also serves as an indicator of maternal health in society. The aim of the study is to examine the risk factors for preterm neonatal mortality at the neonatal intensive care unit (NICU) at Mount Hope Women's Hospital in Trinidad and Tobago (MHWH). In this retrospective study, we included infants (N = 129), born < 37 weeks gestational age, between 1 January and 31 December 2015. Two binary logistic regression models (infant and maternal variables) were constructed to identify predictors of preterm neonatal mortality. Roughly 12% of the infants died after being admitted to the NICU. The binary logistic regression (infant model) had an excellent fit (area under the curve (AUC): 0.904, misclassification rate: 11.7%) whilst the maternal binary logistic model had a fair fit (AUC: 0.698). Birth weight, length of time on the ventilator and obstetric complications proved to significantly influence the odds of preterm neonatal death. The estimated models show that improvement in neonatal as well as maternal variables has direct impact on preterm neonatal mortality.