To the Editor:
We read with great interest the observations by Lenglet et al1 on heated and humidified high-flow nasal cannula (HFNC) oxygen therapy. This technique represents a new alternative to conventional oxygen therapy in the emergency department. The authors' hypothesis was that HFNC is feasible and efficient in patients with acute respiratory failure in the emergency department. This is a potential relevant hypothesis, but, in our view, some concerns must be underlined regarding HFNC in the emergency department.
First, a major factor is the variability in this patient population, which makes it difficult to extrapolate the findings to all patients, and we believe the conclusions should be softened. Although there are some data from pediatric studies,2,3 information is lacking on HFNC versus noninvasive ventilation in adult patients with acute respiratory failure.4 Also, there are some concerns about the optimal FIO2 level to use, since FIO2 could be influenced by the type of mask, the amount of leak, the flow, and the breathing pattern. The results from pediatric studies with regards to the level of pressure applied during HFNC cannot be extrapolated to adults, because of differences in, for example, nasopharynx volume, nasal resistance, and respiratory pattern.5,6
Second, mouth breathing and the degree of mask-face seal affect the pressure applied, assuming that there is certainly loss of pressure through the nostrils or around the mask. When the mouth is open, we assume there is loss of pressure.7 We need more clinical studies to understand the correlation with other parameters such as tachypnea.
Third, hypoventilation is the most frequent situation in emergency admissions for acute respiratory failure, so arterial blood gas sampling would be a more objective evaluation and should be performed.8
Fourth, it is important to evaluate the work load of the respiratory therapists and nurses. To assess the response to conventional ventilation or HFNC we should use validated questionnaires, as was done in other studies, and assess difficulties in implementation.9 In the study by Lenglet et al1 there is no mention of the time required to set up the system.
Fifth, some aspects of hospital organization and healthcare cost deserve attention:
Some patients need transfer to another department or interventions (radiology or specific procedures), and it is important that the therapy be continued. For instance, we need to know what happens when a patient is transferred to the ICU or other wards for tests or procedures.
The number of devices required and the oxygen consumption during HFNC are unknown and would be of interest to compare to conventional oxygen therapy.
Further clinical studies are therefore crucial to rational use of HFNC in the emergency department. Several points should be evaluated, including severity of illness initial HFNC parameters, and whether or not HFNC can decrease ICU admissions or the need for NIV.
Footnotes
The authors have disclosed no conflicts of interest.
- Copyright © 2013 by Daedalus Enterprises
References
The authors respond to: High-Flow Nasal Cannula Oxygen Therapy in the Emergency Department: Welcome, But Selection Should Be the First Step
We thank Drs Esquinas and Martin for their letter regarding high-flow nasal cannula (HFNC) oxygen in the emergency department.1 They raise several concerns related either directly to our work or more generally to the technique.
Regarding our work, we fully agree that initial arterial blood gases (ABGs) are necessary to explore dyspnea in the emergency department. All our patients had initial ABGs. However, due to the observational design of the study, repetition of ABGs was left to the attending physician's discretion. In addition, in the absence of underlying chronic respiratory insufficiency, and if alleviation of respiratory distress is associated with SpO2 > 95–%, we see no reason for immediate and repetitive ABGs, which would not be feasible in emergency departments that admit over 100 patients per day.
The time for set-up of HFNC is approximately 45 seconds longer than for application of a face mask. We do not think that difference has any clinical or practical relevance.
We fully agree with the question regarding hospital organization. However, the most pertinent question is not that of patient transfer to radiology or the ICU, because the HFNC device is easily transported. The relevant question is, can a patient that has been stabilized under HFNC be safely monitored outside the ICU and moved to the ward under HFNC? To the best of our knowledge, there are no data on the subject.
Our bias is that, for a patient admitted for acute hypoxemic respiratory failure, initial monitoring must include continuous respiratory, heart rate, and SpO2 monitoring and regular noninvasive blood pressure measurement. In addition, more aggressive ventilatory support (ie, noninvasive ventilation [NIV] or tracheal intubation) must be readily available. Hence, for these reasons, we believe that patients should be kept under strict surveillance, either in a step-down unit close to the ICU or in the ICU. Obviously, the critical point is the intensity of HFNC support required to stabilize the patient. Once again, there are no data in the literature, but we believe that a requirement of FIO2 > 0.50 and flow of 30 L/min should lead to ICU admission.
Simple arithmetic may answer the question regarding the comparison of oxygen consumption between HFNC and conventional oxygen therapy.
Regarding more general comments on HFNC, we disagree with Martin and Esquinas when they ask for caution regarding extrapolation to all patients. The strength of observational studies lies precisely in the fact that they more accurately reflect routine care of patients than do randomized controlled trails, because they have very limited exclusion criteria. As an example, our studies of HFNC in patients with acute respiratory failure1–3 have included such diverse ARF causes as community-acquired pneumonia, transfusion-related acute lung injury, H1N1 lung infection, drug-induced pneumonitis, COPD exacerbation, cardiogenic pulmonary edema, pulmonary embolism, pneumothorax, sepsis and septic shock, pulmonary contusion, massive pleural effusion, post-extubation respiratory failure, Pneumocystis jiroveci pneumonia, aspiration pneumonia, pneumothorax, and pancreatitis-induced ARDS.
Regarding the comparison with NIV, clearly HFNC is much more easily implemented in the emergency department than is NIV. One of the main advantages of HFNC over NIV is patient tolerance. With HFNC the patient retains speech and oral intake.
Martin and Esquinas raise the point of effective (rather than optimal) FIO2 delivery depending on type of mask and breathing pattern. This is typically a major advantage of HFNC; several studies have clearly shown that FIO2 is constant during HFNC, whatever the flow and breathing frequency: a condition never met with other conventional devices.4,5 In addition, HFNC provides optimal gas conditioning with adequate heat and humidity,6 which provides better comfort than conventional oxygen.7,8
It is inappropriate to extrapolate neonatal data to adults, but that was never our intention. Nonetheless, many of the physiological benefits found in neonates have been reproduced in adults. For example, studies have undisputedly shown that HFNC provides flow-dependent positive airway pressure in adults with mild to moderate respiratory distress.9,10
Regarding tachypnea, breathing frequency has consistently been found improved during HFNC, in comparison with conventional oxygen. All the studies have shown that breathing frequency rapidly decreases under HFNC.11 We have also found that the breathing frequency of patients who required intubation after HFNC initiation was significantly higher than in those who did not, and this difference was observed as early as 15 min.2 Thus, clinicians should be aware that the absence of reduction in breathing frequency after HFNC initiation may predict or at least alert them to the risk of the need for intubation.
Finally, the true question raised by the considerable differences in effective delivered FIO2, alleviation of respiratory distress, tolerance, comfort and quality of gas conditioning (heat and humidity) between conventional and HFNC oxygen is whether or not HFNC should now become the new standard of oxygen delivery. Our contention is that it should.
Footnotes
The author has disclosed a relationship with Fisher & Paykel Healthcare.