High-flow nasal cannula should be first-line therapy for adults with acute hypoxaemic respiratory failure, a new American Thoracic Society guideline recommends.
The clinical practice guideline, published in the American Journal of Respiratory and Critical Care Medicine [link here] is the first to comprehensively address noninvasive respiratory support (NIRS) across the full trajectory of acute respiratory failure, from initial presentation through to extubation.
A multidisciplinary panel, co-chaired by Dr Bruno Ferreyro of the University of Toronto and Dr Neha Goel of the Icahn School of Medicine at Mount Sinai, used the GRADE approach to tackle four key clinical questions covering acute hypoxaemic respiratory failure, acute hypercapnic respiratory failure, preoxygenation before intubation, and post-extubation support.
Strong call for HFNC in hypoxaemic failure
For adults with de novo acute hypoxaemic respiratory failure, the panel issued a strong recommendation for high-flow nasal cannula (HFNC) over standard oxygen therapy, based on moderate-certainty evidence from a pooled analysis of 39 randomised trials involving 2,913 patients.
HFNC cut the risk of needing intubation (RR 0.76, 95% CI 0.64–0.89) and was linked with a modest mortality benefit (RR 0.81, 95% CI 0.67–0.97), though that mortality finding was rated only low certainty.
By contrast, noninvasive ventilation (NIV) delivered via facemask or helmet, and CPAP via either interface, only earned conditional recommendations, on low-certainty evidence.
“The panel determined that the desirable effects of HFNC clearly outweighed the undesirable effects compared with standard oxygen therapy,” the guideline states. “HFNC reduces the need for intubation with minimal harms.”
The authors were careful to note the strong recommendation applies only to HFNC versus standard oxygen, not against NIV. “Evidence comparing HFNC and NIV remains limited and of low certainty, and therefore no recommendation was made for this comparison,” they wrote, adding that the choice should be “individualized based on patient characteristics, clinician expertise, and institutional resources.”
NIV still standard of care for hypercapnic failure
For hypercapnic respiratory failure, the guideline reaffirms NIV via facemask as standard of care, issuing a strong recommendation on moderate-certainty evidence drawn from 43 trials of 5,354 patients, most with COPD exacerbations.
NIV was associated with a near-halving of mortality risk compared with standard oxygen therapy (RR 0.47, 95% CI 0.34–0.65).
HFNC only earns a conditional recommendation as an alternative to NIV, and only in patients with milder disease. The guideline specifies this applies to “patients with acute hypercapnic respiratory failure and mild acidemia (e.g. pH > 7.25) provided that monitoring and escalation to NIV are promptly available.”
The panel flagged a substantial caveat for clinicians tempted to reach for HFNC in this group: across the pooled studies, “one-third of patients initially treated with HFNC required escalation to NIV, underscoring the need for close monitoring and predefined escalation criteria.”
CPAP evidence for hypercapnic failure was judged too dated and heterogeneous to support a recommendation either way, with the panel noting the therapy “is also not commonly used in current practice for hypercapnic respiratory failure, particularly in patients with COPD exacerbation.”
Preoxygenation before intubation
Drawing on 15 trials involving 3,420 patients, the panel made a strong recommendation for using either HFNC or NIV via facemask, rather than standard oxygen, to preoxygenate patients before intubation.
Both cut the risk of peri-intubation hypoxaemia (NIV: RR 0.51, 95% CI 0.39–0.65; HFNC: RR 0.69, 95% CI 0.54–0.88), and head-to-head data suggested NIV “probably reduced hypoxemia compared to HFNC” (RR 0.73, 95% CI 0.55–0.98, moderate certainty). Serious adverse events, including cardiac arrest and haemodynamic instability, were also less frequent with NIV.
Still, the panel stopped short of preferring one modality over the other, citing uncertainty in the comparative data and the practical reality that “initiation of NIV or HFNC should not delay necessary or emergent endotracheal intubation.”
Risk-based approach after extubation
For patients extubated in the ICU after critical illness, the guideline suggests using HFNC or NIV via facemask over standard oxygen therapy, based on moderate-certainty evidence pooled from 53 trials totalling 5,304 patients.
But for those at high risk of extubation failure, including patients aged over 65, those who are obese, hypercapnic, or have undergone prolonged mechanical ventilation, the panel suggests NIV specifically over both standard oxygen and HFNC.
Patients judged very low risk, meanwhile, “could be extubated to standard low flow oxygen or even room air,” the guideline notes.
Device choice matters less than bedside practice
Across all four clinical scenarios, the panel repeatedly stressed that successful implementation hinges less on which device is chosen than on how well it is monitored.
“The effectiveness of these recommendations depends more on consistent and physiological bedside practice than on device choice,” the authors wrote in their discussion. “Institutional efforts should focus on ensuring staff proficiency, streamlined workflows, and well-defined escalation pathways rather than rapid adoption of new technologies.”
The guideline also acknowledges that most of the underlying evidence comes from well-resourced health systems, warning that “the vast majority of the evidence emanates from resource-rich settings” and that broader rollout of NIRS “could widen disparities” if oxygen supply, equipment or trained staff are unevenly distributed.
Two major trials, RENOVATE and the SOHO study, were published after the panel’s evidence synthesis was completed. The authors reviewed both and concluded neither would meaningfully change the pooled estimates or recommendations.
The ATS Quality Improvement and Implementation Committee has flagged the recommendation for NIV in acute hypercapnic respiratory failure as a candidate for future performance measure development.