<p>Background: In acute exacerbations of chronic obstructive pulmonary disease (AECOPD), hypercapnia (raised arterial partial pressure of carbon dioxide, PₐCO₂) is associated with worse clinical outcomes, including death. Nasal high flow (NHF) therapy is a common method of administering oxygen therapy in hypoxic respiratory failure, yet its effect on PₐCO₂ in COPD are uncertain. Aims: To investigate the effects of NHF therapy in people with COPD. Methods: Two randomised controlled trials (RCTs) were undertaken to investigate the effect of NHF on transcutaneous partial pressure of carbon dioxide (PtCO₂) in COPD: one comparing different flow rates (45 L/min, 30 L/min and 15 L/min) to breathing room air as a control in 48 participants with stable COPD, and another comparing NHF against non-invasive ventilation (NIV) in 24 stable COPD patients with chronic hypercapnia. Two further studies tested the feasibility of undertaking multicentre RCTs utilising different applications of NHF in COPD: one in 100 patients (20 in New Zealand, 80 in the United States) using NHF at home for 30 days following hospital discharge to determine patterns of use and rates of hospital readmission, and another in the Emergency Department (ED) to test whether a pre-specified standard protocol for managing acute hypercapnic respiratory failure (AHRF) could be followed. Finally, another RCT compared the tolerability and change in PtCO₂ of a dual NHF/NIV device to a standard NIV device in people with chronic respiratory disorders including COPD. Results: In stable COPD, the mean (95% CI) change in PtCO₂ at 20 min was −0.6 mm Hg (−1.1 to 0.0), P = 0.06; −1.3 mm Hg (−1.9 to 0.8), P < 0.001; and −2.4 mm Hg (−2.9 to −1.8), P < 0.001; for NHF at 15 L/min, 30 L/min and 45 L/min, respectively, compared with room air. In stable COPD with chronic hypercapnia, the mean (SD) reduction in PtCO₂ at 60 min from baseline with NHF was -2.5 mmHg (3.5) compared to -5.3 mm Hg (5.0) with NIV. The PtCO₂ change when analysed across all time points was lower using NIV than with NHF: -2.5mmHg (95% CI -4.5 to -0.5), P=0.016. In the international outpatient feasibility study, mean (SD) NHF use at home following hospital discharge was 1.6 (1.6) hours/day in the 20 NZ participants. The 30-day hospital readmission rate was 10% (95% CI 1.8 to 33.1). Recruitment in the US was unsuccessful, despite a 12-month extension, and no data was available. In the second feasibility study, 120 patients with AECOPD were reviewed over a 4-month period in ED. All 3 patients with AHRF received the standard of care protocol and in one there was significant deviation from the agreed protocol. There was no difference in tolerability between the dual NHF/NIV device and a standard NIV device on a 100mm visual analogue scale: mean difference - 1.3mm (95% CI -7.9 to 5.2, P=0.69), and a lower PtCO2 with the standard NIV device, mean difference -0.61mm Hg (95% CI -1.05 to -0.17, P=0.01). Conclusion: The small, flow-dependent reduction in PtCO₂ observed with the currently available NHF device across the range of flows used in clinical practice indicates that it is safe to use in COPD from a physiological perspective. The reduction on PtCO₂ with NHF was smaller than with NIV, however the difference was of uncertain clinical significance, suggesting NHF represents an alternative therapy for COPD patients with hypercapnia who cannot tolerate NIV, or during breaks from NIV. In NZ, it would be feasible to undertake an RCT using NHF in AECOPD patients discharged from hospital, but not those presenting to ED with AHRF. The recently developed dual NHF/NIV device had similar tolerability as a standard NIV device and there was no clinically significant difference in PtCO2. These findings suggest that NHF represents a therapeutic option for patients with COPD, with the dual NHF/NIV device demonstrating potential as a modality for delivering NHF in COPD patients with AHRF.</p>