Extended CPAP or low-flow nasal cannula for intermittent hypoxaemia in preterm infants: a 24-hour randomised clinical trial.

Objective: Optimal timing of continuous positive airway pressure (CPAP) cessation in preterm infants remains undetermined. We hypothesised that CPAP extension compared with weaning to low-flow nasal cannula (NC) reduces intermittent hypoxaemia (IH) and respiratory instability in preterm infants meeting criteria to discontinue CPAP.
Design: Single-centre randomised clinical trial.
Setting: Level 4 neonatal intensive care unit.
Patients: 36 infants <34 weeks' gestation receiving CPAP≤5 cmH ₂ O and fraction of inspired oxygen (FiO ₂ ) ≤0.30 and meeting respiratory stability criteria.
Interventions: Extended CPAP was compared with weaning to low-flow NC (0.5 L/kg/min with a limit of 1.0 L/min) for 24 hours.
Outcomes: The primary outcome was IH (number of episodes with SpO ₂ <85% lasting ≥10 s). Secondary outcomes included: coefficient of variability of SpO ₂ , proportion of time in various SpO ₂ ranges, episodes (≥10 s) with SpO ₂ <80%, median cerebral and renal oxygenation, median effective FiO ₂ , median transcutaneous carbon dioxide and bradycardia (<100/min for≥10 s).
Results: The median (IQR) episodes of IH per 24-hour period was 20 (6-48) in the CPAP group and 76 (18-101) in the NC group (p=0.03). Infants continued on CPAP had less bradycardia, time with SpO ₂ <91% and <85%, and lower FiO ₂ (all p<0.05). There were no statistically significant differences in IH<80%, median transcutaneous carbon dioxide or median cerebral or renal oxygenation.
Conclusion: In preterm infants meeting respiratory stability criteria for CPAP cessation, extended CPAP decreased IH, bradycardia and other hypoxaemia measures compared with weaning to low-flow NC during the 24-hour intervention.
Trial Registration Number: NCT04792099.
Competing Interests: Competing interests: NA is on the advisory board for Radiometer, Shire and Resbiotic. CPT has applied for a patent with the USPTO for a bradycardia predictor and interrupter unrelated to the current study.
(© Author(s) (or their employer(s)) 2024. No commercial re-use. See rights and permissions. Published by BMJ.)