Miguel Alsina-Casanova ¹ Nerea Brito ² Carla Balcells-Esponera ² Ana Herranz-Barbero ¹ Marta Teresa-Palacio ¹ Aleix Soler-García ² Carmen Agustí ² Guillem Brullas ² Jordi Clotet ² Cristina Carrasco ² Dolors Salvia ¹ Victoria Aldecoa-Bilbao ^1*

• ¹ Hospital Clinic of Barcelona, Barcelona, Spain
• ² Sant Joan de Déu Hospital, Barcelona, Catalonia, Spain

Introduction. Less invasive surfactant administration (LISA) is associated with better respiratory outcomes in preterm infants with respiratory distress syndrome (RDS). However, mechanical ventilation (MV) shortly after the LISA procedure has been related to lower survival. This study aimed to analyze the trends and the main predictors of CPAP failure after LISA. Material and methods. Preterm infants born between 230 and 336 weeks gestational age (GA) in two level III Neonatal Units who received surfactant, were included (2017-2022). Demographic data, lung ultrasound (LUS) score, saturation/fraction of inspired oxygen (SF) ratio, technique, time to surfactant administration, and main neonatal outcomes were collected. Results. Over the study period, 289 inborn preterm infants received surfactant, 174 with the LISA method (60.2%). Patients who received surfactant after intubation in the delivery room (n=56) were more immature and exhibited worse outcomes. Patients who received surfactant via an endotracheal tube (ETT) in the NICU (n=59) had higher LUS scores and a lower SF ratio than those treated with LISA. The LISA method was associated with less death or BPD adjusted for GA, with an aOR=0.37 (95%CI 0.18 – 0.74), p=0.006. CPAP failure after LISA (defined as the need for intubation and MV in the first 72 hours of life) occurred in 38 patients (21.8%), inversely proportional to GA (38.7% in 23-26 weeks, 26.3% in 27-30 weeks, and 7.9% in 30-33 weeks; p<0.001). CPAP failure after LISA was significantly related to death with an aOR=12.0 (95%CI 3.0 – 47.8), p<0.001; and to moderate-severe BPD with an aOR=2.9 (95%CI 1.1 – 8.0), p=0.035; when adjusting for GA. The best predictors of CPAP failure after LISA were GA, intrauterine growth restriction, temperature at admission, SF ratio, and LUS score, with a Nagelkerke’s R2=0.458; p<0.001. The predictive model showed an AUC=0.84 (95% CI 0.75 – 0.93), p<0.001. Conclusions. CPAP failure after LISA is still common in extremely preterm infants leading to an increase in death or disability. Clinicians must acknowledge the main risk factors of CPAP failure to choose wisely the right patient and the best technique. LUS and SF ratio at admission can be useful in deciding.