Increasing continuous positive airway pressure use rates in the delivery room for very preterm infants: a quality improvement initiative.

Background: In spontaneously breathing preterm infants, less invasive strategies, such as continuous positive airway pressure (CPAP) and less invasive surfactant administration, have been increasingly implemented to reduce lung injury. In 2018, our center initiated a noninvasive neonatal resuscitation protocol incorporating these approaches as quality improvement (QI) initiatives. We aimed to evaluate the feasibility, safety, and effectiveness of this protocol by comparing respiratory outcomes before and after its implementation. Methods: We retrospectively reviewed the medical records of 578 infants born at 25 + 0 to 29 + 6 weeks of gestation between 2014 and 2022 at Samsung Medical Center. Infants born in 2018 and those with severe congenital anomalies, delivery room deaths, or outborn status were excluded. The study population was divided into Period 1 (2014-2017, before noninvasive protocol implementation) and Period 2 (2019-2022, after protocol implementation) to assess the impact of QI initiatives on neonatal resuscitation practices. The year 2018 was excluded from the analysis as it represented a transitional period. We analyzed the rate of endotracheal intubation at birth and other respiratory outcomes, such as CPAP failure and bronchopulmonary dysplasia (BPD). Results: The rate of initial intubation was significantly lower in Period 2 than in Period 1 [77.0% vs. 45.9%; adjusted odds ratio (aOR), 0.24; 95% confidence interval (CI), 0.15-0.40; P < 0.001], with declines observed across all gestational groups, particularly among infants ≥26 weeks' gestation. The rate of postnatal steroid use for BPD prevention was also significantly lower in Period 2 (50.0% vs. 15.3%, aOR, 0.12; 95% CI, 0.07-0.21; P < 0.001). In Period 2, among 68 infants initially managed with CPAP during Period 2, 15 (22.1%) experienced CPAP failure within 48 h, and 24 (35.3%) experienced failure at any time during hospitalization. Despite these failures, no significant differences in the incidence of BPD, duration of invasive ventilation, or mortality were observed between the two periods. Conclusions: Our findings suggest that even in extremely preterm infants, a noninvasive resuscitation strategy is feasible, associated with reduced postnatal steroid use, and does not worsen major neonatal outcomes, supporting its use as a viable alternative for those who do not require immediate intubation.