RESUMO
OBJECTIVE: To identify predictors and outcomes of early intubation in preterm infants with respiratory distress, and predictors of need for brief respiratory support (≤1 day). STUDY DESIGN: Secondary analysis of data from a randomized trial comparing nasal high-flow with continuous positive airway pressure as primary respiratory support in preterm infants born at 28-36 weeks of gestation. Intubation was assessed within 72 hours of randomization. RESULTS: There were 564 included infants with a mean (SD) gestational age of 32.0 (2.2) weeks and birth weight 1744 (589) g; 76 infants (13.5%) received early intubation. On multivariable analysis, lower gestational age and higher pre-randomization fraction of inspired oxygen (FiO2) predicted intubation. A test based on gestational age of <30 weeks and an FiO2 of ≥0.30 produced a likelihood ratio of 9.1. Intubation was associated with prolonged duration of respiratory support and supplemental oxygen, with pneumothorax and nasal trauma, and in infants born at <32 weeks of gestational, with bronchopulmonary dysplasia and patent ductus arteriosus requiring treatment. Greater gestational age and lower FiO2 predicted the need for ≤1 day of respiratory support. A test based on a gestational age of ≥34 weeks and an FiO2 of 0.21 produced a likelihood ratio of 4.7. CONCLUSIONS: In preterm infants 28-36 week of gestation receiving primary noninvasive respiratory support, lower gestational age, and higher FiO2 predicted need for intubation within 72 hours. Intubation was associated with adverse respiratory outcomes. Greater gestational age and lower FiO2 predicted need for ≤1 day of respiratory support. It may be reasonable to defer the use of respiratory support in more mature infants with low FiO2 requirements. TRIAL REGISTRATION AUSTRALIAN NEW ZEALAND CLINICAL TRIALS REGISTRY: ACTRN12613000303741.