Despite the paucity of evidence, the practice of weaning nasal continuous positive airway pressure (NCPAP) is widespread. However, the most clinically effective non-invasive ventilatory support strategy remains to be determined. We compared the outcome of very premature infants with respiratory distress syndrome treated with a combination of NCPAP and heated humidified high-flow nasal cannula (HHFNC) versus NCPAP and low-flow nasal cannula (LFNC). Between 2004 and 2008, patients ≤28 weeks of gestation and <1,250 g of birth weight were treated with NCPAP + HHFNC or NCPAP + LFNC. Their respiratory and non-respiratory outcome including cost-effectiveness was compared after matching for antenatal steroid doses, mode of delivery, birth plurality, gestational age, birth weight, gender, surfactant doses, length of mechanical ventilation and clinical risk index for babies-II (CRIB-II) score. Thirty-nine infants received HHFNC + NCPAP, and 40 received NCPAP + LFNC. Median gestational age and birth weight were 27 weeks and 930 g and 27 weeks and 980 g, respectively. The total number of NCPAP days was significantly reduced by 50 % in the HHFNC group. Thirteen percent of the patients on NCPAP suffered from nasal bridge lesions compared to none on HHFNC. Respiratory and non-respiratory outcome was not significantly different otherwise. Combination of NCPAP and HHFNC reduced costs by 33 %.
Conclusions: HHFNC shortens NCPAP time without increasing overall length of non-invasive respiratory support in very preterm infants. Unlike NCPAP, HHFNC does not seem to increase the risk of nasal trauma and appears to improve cost-effectiveness whilst producing otherwise equal respiratory and non-respiratory outcomes.