Predictors of in-hospital mortality among critically ill neonates admitted to a tertiary NICU in Ecuador.

BACKGROUND

Neonatal mortality remains a major challenge in low- and middle-income countries, where locally derived evidence from neonatal intensive care units is limited. We aimed to identify factors associated with in-hospital mortality among critically ill neonates admitted to a tertiary NICU in Quito, Ecuador, and to evaluate an internally validated mortality prediction model.

METHODS

We conducted a retrospective cohort study of inborn neonates aged ≤ 28 days admitted to the NICU of Isidro Ayora Gynecology and Obstetrics Hospital from January 1 to December 31, 2022. Maternal, perinatal, neonatal, early clinical severity, laboratory, respiratory support, clinical course, and resource utilization variables were extracted from medical records. The primary outcome was in-hospital mortality. Bivariate analyses and multivariable logistic regression were performed. Model performance was assessed using discrimination, calibration, bootstrap internal validation, and decision curve analysis.

RESULTS

Among 220 critically ill neonates, 87 (39.5%) died. Small-for-gestational-age status was independently associated with mortality after adjustment for birth weight category, gestational age, sex, congenital malformations, and antenatal corticosteroid exposure (adjusted OR, 2.57; 95% CI, 1.29-5.14; P=.01). Birth weight < 1400 g was associated with greater clinical severity and resource use but not independently with mortality. Early hemodynamic instability, metabolic acidosis, higher lactate, higher FiO₂, thrombocytopenia, and higher oxygenation index were more frequent among non-survivors. The final model showed good discrimination (apparent AUC, 0.870; optimism-corrected AUC, 0.854) and acceptable calibration. Among survivors, each additional day of invasive mechanical ventilation was associated with longer NICU stay.

CONCLUSION

In this tertiary Ecuadorian NICU cohort, small-for-gestational-age status was independently associated with mortality. These findings support fetal growth assessment, early severity evaluation, and locally validated neonatal risk stratification.