Use of evidence based practices to improve survival without severe morbidity for very preterm infants:results from the EPICE population based cohort

OBJECTIVES: To evaluate the implementation of four high evidence practices for the care of very preterm infants to assess their use and impact in routine clinical practice and whether they constitute a driver for reducing mortality and neonatal morbidity.DESIGN: Prospective multinational population based observational study.SETTING: 19 regions from 11 European countries covering 850 000 annual births participating in the EPICE (Effective Perinatal Intensive Care in Europe for very preterm births) project.PARTICIPANTS: 7336 infants born between 24+0 and 31+6 weeks' gestation in 2011/12 without serious congenital anomalies and surviving to neonatal admission.MAIN OUTCOME MEASURES: Combined use of four evidence based practices for infants born before 28 weeks' gestation using an "all or none" approach: delivery in a maternity unit with appropriate level of neonatal care; administration of antenatal corticosteroids; prevention of hypothermia (temperature on admission to neonatal unit ≥36°C); surfactant used within two hours of birth or early nasal continuous positive airway pressure. Infant outcomes were in-hospital mortality, severe neonatal morbidity at discharge, and a composite measure of death or severe morbidity, or both. We modelled associations using risk ratios, with propensity score weighting to account for potential confounding bias. Analyses were adjusted for clustering within delivery hospital.RESULTS: Only 58.3% (n=4275) of infants received all evidence based practices for which they were eligible. Infants with low gestational age, growth restriction, low Apgar scores, and who were born on the day of maternal admission to hospital were less likely to receive evidence based care. After adjustment, evidence based care was associated with lower in-hospital mortality (risk ratio 0.72, 95% confidence interval 0.60 to 0.87) and in-hospital mortality or severe morbidity, or both (0.82, 0.73 to 0.92), corresponding to an estim