Objectives: To determine mortality rate-dependent variations in the timing and causes of death, and to subsequently identify the clinical factors associated with decreased mortality in extremely preterm infants born at 23-24 weeks' gestation.
Design: A retrospective cohort study.
Setting: Korean Neonatal Network registry that includes all level greater than or equal to 3 neonatal ICUs in Korea.
Patients: Eligible, actively treated infants born at 23-24 weeks' gestation (n = 574) from January 2014 to December 2016 were arbitrarily categorized based on institutional mortality rates of less than or equal to 50% (group I, n = 381) and greater than 50% (group II, n = 193). The primary outcome was mortality before discharge and the timing and causes of death according to the mortality rate.
Interventions: None.
Measurements and main results: The overall mortality rate was significantly lower in group I (40.7%) than in group II (79.3%). Regarding causes of death, mortalities due to cardiorespiratory, infectious, and gastrointestinal causes were significantly lower in group I than in group II. Mortality rates were significantly lower in group I, including all the subgroups that were categorized according to the timing of death, than in group II. The multivariate analyses showed that antenatal corticosteroid use, absence of oligohydramnios, birth weight, and body temperature at admission to the neonatal ICU were significantly associated with reduced mortality.
Conclusions: The reduced mortality rate among the infants born at 23-24 weeks' gestation was attributable to decreased mortality ascribed to cardiorespiratory, infectious, and gastrointestinal causes, and it was associated with antenatal steroid use and body temperature at admission to the neonatal ICU.