Seriously, Should We Be Treating Severe ARDS With High-Flow Nasal Cannula Oxygen?

To the Editor: We have read with interest the RESPIRATORY CARE article entitled “Use of high-flow nasal cannula oxygen therapy in subjects with ARDS: a 1-year observational study”.1 We are surprised that, compared with an overall mortality of 29%, the mortality in the group treated with high-flow nasal cannula (HFNC) oxygen therapy and subsequently needing intubation was 50%. This is close to the value determined by Antonelli et al2 in 2007: 56% mortality in subjects treated with noninvasive ventilation (NIV) and finally intubated due to NIV failure. Because there was a possible delay in the intubation of these subjects the result was a higher mortality rate. In addition, mortality could have been prevented if, instead of being treated with NIV, patients with ARDS were initially intubated and treated with the open-lung strategy with high PEEP.3 The current mortality due to ARDS varies between 20 and 60% depending on many factors, but there is a clear relationship that has been established between the level of PEEP, the PaO2/FIO2 that is reached with this PEEP, and mortality.4 Thus, in patients intubated and ventilated with a PEEP of 10 cm H2O, if the PaO2/FIO2 is 150, the mortality rate is 60.3%, which is very similar to the rate found by Antonelli et al2 in subjects with delayed intubation. InARDS, theventilationstrategy isaimed at reducing the intrapulmonary shunt with the use of continuous distending pressure. Due to the high rate of associated failure and secondary mortality, there are some physicians who advise against the continuation of NIV in patients with ARDS if, after 1 h of treatment, oxygenation has not improved to a PaO2/FIO2 of 175 2 (the intrapulmonary shunt when PaO2/FIO2 175 is 35% 5). In recent clinical trials involving hypoxemic respiratory failure in the pediatric population, HFNC was not shown to have better results than standard low-flow oxygen therapy,6 and both therapies had a higher mortality rate than CPAP therapy.7 In mild hypoxemic situations, although NIV produces better oxygenation, HFNC is subjectively better tolerated, and it can be considered as a reasonable alternative therapy.8 However, we disagree with the use of HFNC in moderate or severe ARDS (PaO2/ FIO2 of 200). ARDS is a clinical condition with high mortality. If the decision to use NIV is made, it should be started with CPAP or bi-level positive airway pressure. However, if an improvement in oxygenation (PaO2/FIO2 of 175) is not obtained after 1 h of NIV, the patient should be intubated to improve the level of recruitment and to minimize the intrapulmonary shunt. So the question now is, are we sure that using HFNC initially, instead of conventional mechanical ventilation in patients with severe ARDS, does not increase mortality? In our opinion, the answer is no. If, as patients, we happened to have severe ARDS, we would definitely choose conventional mechanical ventilation from the very beginning.