LAUSR

We read with concern the article by Golmohamad et al. who reported on the ‘safety and efficacy of high-flow nasal cannula (HFNC) therapy in acute hypercapnic respiratory failure’. The study results leave us questioning whether either safety or efficacy has been demonstrated and we would dispute the suggestion that they represent compelling pilot data to undertake a prospective trial. Golmohamad et al. retrospectively compared outcomes from heterogeneous groups in their single-centre study. Participants had a range of disorders resulting in acute hypercapnic respiratory failure, but were managed with two different approaches – non-invasive ventilation (NIV) or HFNC. Those receiving NIV were more unwell based on blood gas derangements (lower pH and higher arterial partial pressure of carbon dioxide (PaCO2)), but no further description of clinical reasoning was provided as to treatment choice. Concerningly, one in four treated with HFNC failed therapy and required NIV rescue. Therefore, the authors’ conclusion that ‘in real-world conditions, HFNC was safe and effective in normalising blood gas parameters for most patients with mild acute hypercapnic respiratory failure’ is not supported. In our opinion, their data indicate that most patients with acute hypercapnic respiratory failure needed NIV and a large proportion would have benefited significantly if they had received it earlier in their admission. It is worth reflecting on the historical efficacy of managing acute hypercapnia due to acute exacerbations of chronic obstructive pulmonary disease (AECOPD) without NIV. Controlled oxygen, bronchodilators, corticosteroids, antibiotics and occasionally diuretics can be effective, but better outcomes can be achieved with the addition of acute NIV. Plant et al. demonstrated that early NIV delivered outside an intensive care unit setting for mild– moderate acidosis due to AECOPD was associated with an in-hospital mortality rate of 10% compared with standard care whose mortality rate was 20%. So, although many of those with AECOPD and acute hypercapnia do survive to discharge without NIV, choosing an alternative initial therapy would be ignoring the weight of evidence that confirms that NIV reduces length of stay, intubation and mortality rates. Therefore, we would be concerned if the results reported by Golmohamad et al. produced any shift in practice towards the use of HFNC for acute hypercapnic respiratory failure. This concern exists for those with AECOPD but also for people with restrictive chest wall disorders, neuromuscular disorders or obesity-related respiratory failure, all of whom are likely to fare incredibly poorly if they are managed acutely with HFNC instead of NIV. Rather than looking for new ways to use HFNC, greater efforts should be made to deliver better acute NIV more often. In the UK, annual audits of ward-based NIV have reversed concerning trends in the mismanagement of acute NIV through a clear quality improvement framework. The strategy emphasises improved patient selection, rapid identification of acidosis and hypercapnia with arterial blood gases, faster commencement of acute NIV once it is indicated, better staff training, clear escalation protocols and closer monitoring. Improvements in the quality of acute NIV should be prioritised as it is likely to be a much safer and more efficacious option than any shift towards HFNC for acute hypercapnic respiratory failure.