High-flow nasal cannulae, bronchopulmonary dysplasia and retinopathy of prematurity

Aims To determine if HFNC use was associated with changes in incidence of BPD and ROP. Methods This retrospective study examined premature infants (<30 weeks GA or <1500g) in a tertiary neonatal unit from 2010-2016. Patients were compared before and after introduction of HFNC. Further analysis of high-risk infants (<28 weeks GA or <750g or ventilated) compared those who received HFNC to those who did not across the whole period. Primary outcomes were incidence of BPD and ROP requiring surgery. Results Incidence of BPD rose following the introduction of HFNC (82/232 (35.3%) after vs 33/251 (13.1%) before, p<0.001). On multivariate analysis, the chance of developing BPD after HFNC introduction remained higher (OR 4.353, 95% CI 2.546-7.443). More infants received surgery for ROP following HFNC introduction (0/214 vs 11/205 (5.4%), p=<0.001). In the second analysis, the rate of BPD was higher in those who received HFNC (90/132 (68.1%) vs 33/153 (21.6%), p<0.001). Receiving HFNC demonstrated higher chance of BPD in multivariate analysis (OR 7.802, 95% CI 4.223-14.423). Rate of ROP surgery was higher in those who received HFNC (0/153 vs 13/134 (9.7%), p<0.001). Conclusions In this study, use of HFNC was associated with significantly increased risk of adverse outcomes.