[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
Purchase Options:
[Skip to Content Landing]
Views 7,465
Citations 0
October 24, 2018

High-Flow Nasal Oxygen—The Pendulum Continues to Swing in the Assessment of Critical Care Technology

Author Affiliations
  • 1Department of Medicine, University of Chicago, Chicago, Illinois
  • 2Section of Pulmonary and Critical Care Medicine, University of Chicago, Chicago, Illinois
  • 3University of Chicago, Chicago, Illinois
JAMA. Published online October 24, 2018. doi:10.1001/jama.2018.14287

Standard oxygen by mask or nasal prongs has been the first-line therapy for patients with acute hypoxemic respiratory failure (AHRF), followed by intubation to provide invasive mechanical ventilation for patients for whom this approach has failed. Although intubation and subsequent invasive mechanical ventilation can be lifesaving, these procedures are associated with many complications1 and some patients with comorbidities—for example, those who are immunosuppressed—have disproportionately high morbidity and mortality.2 Two technologies have been developed to bridge the therapy gap between standard oxygen therapy and invasive mechanical ventilation: noninvasive ventilation (NIV) and high-flow nasal oxygen therapy. Although there is controversy about how these technologies fit in the management of AHRF, they share a similar proliferation based on early enthusiasm followed by widespread adoption and over time a tempering of expectations related to ongoing evaluation in randomized clinical trials.

Although NIV was first introduced in the 1940s, its popularity for the care of immunosuppressed patients with AHRF increased when initial clinical trials reported substantial improvements in mortality and a reduction in rates of endotracheal intubation.3,4 These data and an observational study of 1302 immunosuppressed patients were the basis for a conditional recommendation for the use of NIV in immunocompromised patients with AHRF before intubation in the current European Respiratory Society/American Thoracic Society guidelines.5 However, a recent multicenter, randomized trial by Lemiale et al6 of 374 immunosuppressed patients showed that early NIV compared with standard oxygen therapy was not associated with clinical benefits. In addition, a post hoc analysis of the FLORALI trial, comparing high-flow nasal oxygen therapy with NIV and standard oxygen therapy in AHRF, suggested that NIV might be associated with an increased risk of intubation and mortality in this subgroup of patients with AHRF.7

Given the pendulum swing in optimism for NIV, could there be a role for high-flow nasal oxygen therapy in immunocompromised patients? This technology has been widely adopted, and its popularity has been driven by early positive studies, the improvement in physiologic parameters seen during its use (particularly an increase in the ratio of Pao2 to fraction of inspired oxygen [Fio2]), and a general ease of application.8,9 However, robust studies clarifying the niche this technology best serves have been lacking, especially in the immunosuppressed patient population.

In this issue of JAMA, Azoulay and colleagues10 address this question. In this multicenter trial, the authors recruited 776 immunosuppressed patients with AHRF and randomized them to receive high-flow nasal oxygen therapy vs standard oxygen therapy. Based on their calculated Pao2:Fio2 ratio both at and 6 hours after randomization and the high mortality in both the intervention and the control groups, the population studied was appropriate in regard to degree of hypoxemia and severity of illness to warrant consideration of innovative therapies beyond routine initial oxygen therapy to improve outcome. However, no significant benefit from use of high-flow oxygen therapy was seen. Intubation rates were similar in both groups, 150 of 388 (38.7%) with high-flow oxygen therapy and 170 of 388 (43.8%) with standard oxygen therapy. Similarly, 28-day mortality was not significantly different between groups—138 of 388 (35.6%) with high-flow oxygen therapy and 140 of 388 (36.1%) with standard oxygen therapy.

There may have been some adverse effects of high-flow nasal oxygen therapy in the trial. The authors point out that patients who received high-flow oxygen therapy had a longer intensive care unit (ICU) stay when compared with patients receiving standard oxygen therapy (8 days vs 6 days), although the difference was not statistically significant (P = .07). This observation is common in clinical practice, whereby patients with AHRF who receive high-flow oxygen therapy are considered to need a high-maintenance, high-cost admission to an ICU until they can be transitioned to standard oxygen therapy. Furthermore, because there are no precise guidelines for weaning from high-flow therapy, its use may lead to increased and perhaps unnecessary use of hospital and critical care resources.

The medical community’s craving for innovation often fuels overzealous enthusiasm for positive results of interventions in preliminary studies that are subsequently contradicted when larger, multicenter trials are undertaken.11 One reason for early enthusiasm is that physicians do not want to withhold potentially beneficial therapies from patients. This is especially true in critical care when the intervention is perceived to have a pathophysiologic rationale. However, once a technology has been adopted, it is difficult to de-adopt, even if later, more robust evidence suggests that its continued use is unjustified.16 Even when trials have negative results, researchers and clinicians often seek to find subgroups that may have some benefit (such as with trials of colloids in shock or of activated protein C in severe sepsis) so that innovation is not wasted. However, if therapies are posited to have a role in important subgroups, it is important that such a role be demonstrated with adequate rigor in prospective clinical trials.12,13

It is conceivable that one reason there has been such an embrace of high-flow nasal oxygen therapy is publication bias. Researchers may not submit their negative studies because they perceive their results are uninteresting, or journal priorities and the agenda of funding groups may influence the dissemination of information from completed clinical trials by limiting publication.14 However, trials that fail to demonstrate positive effects of new technologies or therapies often have clinical utility. The trial by Azoulay et al, despite its negative findings, helps clarify the application of high-flow oxygen therapy in the immunosuppressed patient population. It is important to publish high-quality, negative randomized clinical trials to prevent excessive application of therapies that are not beneficial and, once popularized, may take years to find their proper application.

Given the available evidence, the important clinical question is in which patients with AHRF should high-flow nasal oxygen therapy be used? Based on the post hoc subgroup analysis in the FLORALI trial, high-flow oxygen therapy used in patients with severe hypoxemia (Pao2:Fio2 ratio ≤200 mm Hg) was associated with a reduced rate of intubation, which likely drove the mortality benefit of high-flow oxygen therapy compared with NIV and standard oxygen therapy.15 However, the patients in the current study by Azoulay et al had severely impaired oxygenation yet did not benefit. Thus, based on current information, high-flow oxygen therapy should not be considered a preferred therapy for immunosuppressed patients with AHRF, and additional studies including assessment of other technologies to avoid invasive ventilation are clearly needed.

Back to top
Article Information

Corresponding Author: Jesse B. Hall, MD, Section of Pulmonary and Critical Care Medicine, University of Chicago, 5841 S Maryland Ave, MC 6026, Chicago, IL 60637 (jhall@medicine.bsd.uchicago.edu).

Published Online: October 24, 2018. doi:10.1001/jama.2018.14287

Conflict of Interest Disclosures: Dr Patel reported grants from Parker B. Francis Foundation. No other disclosures were reported.

Dos Santos  C, Heunks  L, Wunsch  H.  Update in critical care 2016.  Am J Respir Crit Care Med. 2017;196(1):11-17. doi:10.1164/rccm.201701-0164UPPubMedGoogle ScholarCrossref
Cortegiani  A, Madotto  F, Gregoretti  C,  et al; LUNG SAFE Investigators, ESICM Trials Group.  Immunocompromised patients with acute respiratory distress syndrome: secondary analysis of the LUNG SAFE database.  Crit Care. 2018;22(1):157. doi:10.1186/s13054-018-2079-9PubMedGoogle ScholarCrossref
Hilbert  G, Gruson  D, Vargas  F,  et al.  Noninvasive ventilation in immunosuppressed patients with pulmonary infiltrates, fever, and acute respiratory failure.  N Engl J Med. 2001;344(7):481-487. doi:10.1056/NEJM200102153440703PubMedGoogle ScholarCrossref
Antonelli  M, Conti  G, Bufi  M,  et al.  Noninvasive ventilation for treatment of acute respiratory failure in patients undergoing solid organ transplantation: a randomized trial.  JAMA. 2000;283(2):235-241. doi:10.1001/jama.283.2.235PubMedGoogle ScholarCrossref
Rochwerg  B, Brochard  L, Elliott  MW,  et al.  Official ERS/ATS clinical practice guidelines: noninvasive ventilation for acute respiratory failure.  Eur Respir J. 2017;50(2):1602426. doi:10.1183/13993003.02426-2016PubMedGoogle ScholarCrossref
Lemiale  V, Mokart  D, Resche-Rigon  M,  et al; Groupe de Recherche en Réanimation Respiratoire du patient d’Onco-Hématologie (GRRR-OH).  Effect of noninvasive ventilation vs oxygen therapy on mortality among immunocompromised patients with acute respiratory failure: a randomized clinical trial.  JAMA. 2015;314(16):1711-1719. doi:10.1001/jama.2015.12402PubMedGoogle ScholarCrossref
Frat  JP, Ragot  S, Girault  C,  et al; REVA Network.  Effect of non-invasive oxygenation strategies in immunocompromised patients with severe acute respiratory failure: a post-hoc analysis of a randomised trial.  Lancet Respir Med. 2016;4(8):646-652. doi:10.1016/S2213-2600(16)30093-5PubMedGoogle ScholarCrossref
Frat  JP, Brugiere  B, Ragot  S,  et al.  Sequential application of oxygen therapy via high-flow nasal cannula and noninvasive ventilation in acute respiratory failure: an observational pilot study.  Respir Care. 2015;60(2):170-178. doi:10.4187/respcare.03075PubMedGoogle ScholarCrossref
Huang  HB, Peng  JM, Weng  L, Liu  GY, Du  B.  High-flow oxygen therapy in immunocompromised patients with acute respiratory failure: a review and meta-analysis.  J Crit Care. 2018;43:300-305. doi:10.1016/j.jcrc.2017.09.176PubMedGoogle ScholarCrossref
Azoulay  E, Lemiale  V, Mokart  D,  et al.  Effect of high-flow nasal oxygen vs standard oxygen on 28-day mortality in immunocompromised patients with acute respiratory failure: the HIGH randomized clinical trial  [published online October 24, 2018].  JAMA. doi:10.1001/jama.2018.14282Google Scholar
Ioannidis  JPA.  Contradicted and initially stronger effects in highly cited clinical research.  JAMA. 2005;294(2):218-228. doi:10.1001/jama.294.2.218PubMedGoogle ScholarCrossref
Caironi  P, Tognoni  G, Masson  S,  et al; ALBIOS Study Investigators.  Albumin replacement in patients with severe sepsis or septic shock.  N Engl J Med. 2014;370(15):1412-1421. doi:10.1056/NEJMoa1305727PubMedGoogle ScholarCrossref
Abraham  E, Laterre  PF, Garg  R,  et al; Administration of Drotrecogin Alfa (Activated) in Early Stage Severe Sepsis (ADDRESS) Study Group.  Drotrecogin alfa (activated) for adults with severe sepsis and a low risk of death.  N Engl J Med. 2005;353(13):1332-1341. doi:10.1056/NEJMoa050935PubMedGoogle ScholarCrossref
Montori  VM, Smieja  M, Guyatt  GH.  Publication bias: a brief review for clinicians.  Mayo Clin Proc. 2000;75(12):1284-1288. doi:10.4065/75.12.1284PubMedGoogle ScholarCrossref
Frat  JP, Thille  AW, Mercat  A,  et al; FLORALI Study Group; REVA Network.  High-flow oxygen through nasal cannula in acute hypoxemic respiratory failure.  N Engl J Med. 2015;372(23):2185-2196. doi:10.1056/NEJMoa1503326PubMedGoogle ScholarCrossref
Niven  DJ, McCormick  TJ, Straus  SE,  et al.  Reproducibility of clinical research in critical care: a scoping review.  BMC Med. 2018;16(1):26. doi:10.1186/s12916-018-1018-6PubMedGoogle ScholarCrossref