急性心源性肺水肿中高流量鼻导管与无创通气的对比：随流而动？

原文图文 · 图 32 张

REBEL Rundown

Key Points

HFNC vs. NIV: High-flow nasal cannula (HFNC) performed similarly to non-invasive ventilation (NIV) in reducing respiratory rates at 120 minutes in patients with acute cardiogenic pulmonary edema (ACPE).

Study Groups: No significant differences were found between the two groups regarding dyspnea scores or arterial blood gas parameters during the two-hour follow-up period.

Patient Comfort: HFNC may offer a more tolerable and comfortable experience for patients compared to the masks associated with NIV, with similar clinical effect

Treatment Failures: 29 patients required a treatment switch, 16 moved from HFNC to NIV and 13 from NIV to HFNC.

Alternative Strategy: HFNC appears to be a viable alternative to NIV in the initial management of moderate ACPE, particularly for those patients who do not tolerate positive pressure masks.

Introduction

Acute cardiogenic pulmonary edema (ACPE) is a common cause for emergency department (ED) visits, often resulting in hypoxemia, hypercarbia and respiratory distress. While non-invasive ventilation (NIV) remains the established first-line treatment for its ability to provide positive pressure and reduce cardiac workload, patient tolerance is frequently a limiting factor. High-flow nasal cannula (HFNC) has emerged as a possible alternative due to its ability to deliver humidified oxygen at high flow rates with modest PEEP. REBEL EM has previously explored the mechanisms of High-Flow Nasal Oxygen in Respiratory Failure and compared oxygenation strategies like BPAP vs AVAPS, but direct comparisons in the ACPE population remain limited. The authors of the following randomized trial sought to investigate if HFNC is as effective as NIV in improving respiratory outcomes for ACPE patients.

Paper

Altunbas E, et al. Is high-flow nasal oxygen as effective as non-invasive ventilation in acute cardiogenic pulmonary Edema? Am J Emerg Med. 2025. PMID: 40848477

PREVIOUSLY COVERED ON REBEL EM:

High-Flow Nasal Oxygen in Respiratory Failure- How it works

BPAP vs AVAPS for Hypercapnic Respiratory Failure in the ED

Is All Oxygen Created Equal?

RENOVATE Trial

What They Did

CLINICAL QUESTION

Is HFNC as effective as NIV in reducing respiratory rate and improving respiratory distress within 2 hours in ACPE?

STUDY DESIGN

Prospective, single-center, randomized superiority trial (1:1) of 178 patients (selected from 1,376 screened) conducted in an academic ED Turkey, between July 2023 and April 2024, with Intention-to-Treat (ITT) and Per-Protocol (PP) analyses.

Randomization: Computer-generated sequence with concealed allocation (sealed envelopes).

Blinding: Treating clinicians and patients were unblinded; outcome assessors and data analysts were blinded.

Treatment Duration: Patients received the assigned intervention (HFNC or NIV) for a minimum of 2 hours.

Dyspnea Assessment: Utilized the Modified Borg Dyspnea Scale (MBDS), a subjective scoring system for shortness of breath ranging from 0 (nothing at all) to 10 (maximal). A change of ≥1 point was defined as clinically significant

Standard Care: All patients received standard medical therapy for ACPE, including intravenous diuretics and nitrates according to established guidelines.

De-Escalation Protocol: Switching to standard nasal cannula was allowed if clinical endpoints were met (RR <24 breaths/min, SpO₂ >92%, and decreased work of breathing).

Treatment Failure/Switching: Switching to alternative methods (e.g., NIV or intubation) was allowed for intolerance or clinical deterioration (defined as persistent/worsening respiratory symptoms, increased RR, elevated CO₂, or decreased SpO₂).

POPULATION

Inclusion Criteria:

Inclusion Criteria: Adults ≥18 years old presenting to the ED with suspected ACPE and meeting the following criteria:

Respiratory Rate >24/min

SpO₂ <92% on room air

Increased work of breathing (accessory muscle use or paradoxical breathing)

Bilateral rales on physical exam

Radiographic or sonographic evidence of pulmonary congestion, including: Pulmonary venous congestion, cardiomegaly, or interstitial edema on chest radiography

Sonographic interstitial syndrome (multiple, bilateral, and homogeneously distributed B-lines on lung ultrasound)

Exclusion Criteria:

Need for immediate endotracheal intubation.

Hemodynamic instability (MAP ≤65 mmHg or requirement for vasopressors).

Altered mental status (GCS ≤13).

ST-Elevation Myocardial Infarction (STEMI).

End-Stage Renal Disease (ESRD).

Pregnancy.

Contraindications to NIV or HFNC (e.g., facial trauma, vomiting).

INTERVENTION & COMPARATOR

Intervention:

HFNC at 60 L/min, FiO₂ titrated to SpO₂ ≥92%.

Comparator:

NIV via oro-nasal mask in Continuous Positive Airway Pressure mode (CPAP) with PEEP set between 5–10 cmH₂O and FiO₂ titrated to SpO₂ ≥92%

Both groups received standard ACS management (hydration, pain control, antibiotics, transfusion per predefined criteria, oxygen/respiratory support)

OUTCOMES

Primary OutcomeSecondary Outcomes

Change in RR at 120 minutes

Changes in vital signs

Arterial blood gases (pH, pCO2 and lactate)

Modified Borg Dyspnea Scale (MBDS) within 120 minute

Treatment switches

Hospitalization rates

All-cause death

At 6 months:

Hospital readmissions (all-cause and ACS-specific)

Thrombotic events (DVT, pulmonary artery thrombosis)

All-cause death

Results

Baseline Characteristics

Baseline Characteristics

Critical Results

Strengths

Prospective randomized controlled design: The trial was conducted as a prospective, randomized superiority study comparing HFNC and NIV head-to-head in adults with acute cardiogenic pulmonary edema (ACPE). Randomization successfully created balanced baseline characteristics between groups (e.g., similar age, comorbidities, and initial vital signs), ensuring the cohorts started with similar disease severity before interventions were applied.

Consecutive sampling: The study screened all consecutive patients presenting to the ED rather than a convenience sample. This minimizes selection bias by ensuring that investigators did not “cherry-pick” participants, making the cohort more representative of the general ACPE population.

Clinically relevant comparator and question: NIV is guideline-recommended first-line noninvasive support in ACPE, so directly comparing HFNC against NIV (rather than against conventional oxygen) answers a practical bedside question: Can HFNC safely substitute for NIV in this setting?

Robust analysis strategy (ITT + per-protocol): Reporting both intention-to-treat and per-protocol analyses allows you to see the effect of randomization (real-world practice including cross-over) and the effect in those who actually received the assigned therapy, which increases confidence that the “no difference” finding isn’t purely an artifact of protocol deviations.

Consistent physiologic outcome assessment: The study repeatedly measured respiratory rate, vital signs, ABG parameters, and dyspnea scores at 30, 60, and 120 minutes. This dense physiologic sampling provides a detailed picture of how each modality affects short-term respiratory failure physiology, rather than relying on a single time point.

Clinically interpretable negative result: Demonstrating no meaningful difference between HFNC and NIV in improving RR and other physiologic markers supports HFNC as a non-inferior alternative in many ACPE patients—particularly useful when NIV is poorly tolerated or resources are limited.

Limitations

Single-center study: All patients were enrolled in one Turkish academic ED. Practice patterns, patient demographics, and heart-failure management strategies may differ in other hospitals or countries, which limits generalizability of the findings.

Short observation window: Data collection stopped at 120 minutes. Clinical deteriorations (e.g., respiratory fatigue, late need for intubation) occurring after this window would be missed, meaning the study cannot confirm the durability of the treatment effect beyond the first two hours. It’s possible after 120 minutes there would have been statistically or clinically significant changes in patient outcomes.

Lack of patient-centered outcomes: The study focused on a surrogate endpoint (respiratory rate) and did not collect data on hospital length of stay, re-admission rates, or 30-day mortality. Therefore, we do not know if the initial physiologic improvement translates to sustained recovery or if it merely delays failure.

Modest sample size for treatment-failure outcomes: With 178 randomized patients, the trial has reasonable power for a physiologic endpoint but is likely underpowered to detect small–moderate differences in rare outcomes (e.g., intubation, in-hospital death). A true difference could exist but remain statistically undetected.

Open-label design: Neither clinicians nor patients were blinded to treatment assignment. Knowledge of which device was being used could have influenced co-interventions (diuretics, vasodilators, decisions to escalate or persist with therapy), introducing performance bias.

Exclusion after randomization due to resource failure: Two patients were randomized to NIV but then excluded because the hospital physically ran out of ventilator sets. This violates the Intention-to-Treat principle and introduces selection bias, though it highlights the logistical and resource constraints of NIV compared to HFNC.

Lack of standardized medical therapy protocol: Although the study states patients received “standard” intravenous diuretics and nitrates, there was no strict protocol for dosing, timing, or escalation. Variability in the aggressiveness of medical management (e.g., high-dose nitroglycerin vs. standard dosing) could act as a major confounder, making it difficult to determine if patient improvement was driven by the oxygen device or the pharmacotherapy.

Incomplete data for key secondary outcomes: ABG-based measures (PCO₂, lactate, MBDS) were analyzed only in subsets of patients with interpretable data. Reducing statistical power and raising the possibility that those with missing data differed systematically from those included, which could lead to attrition bias. (This is hinted at by the footnotes indicating analyses in smaller numbers than the full cohort.)

Limited applicability to the sickest ACPE patients: As with most NIV/HFNC trials, patients requiring immediate intubation or with significant hemodynamic instability were excluded, meaning the results are most applicable to moderately ill ACPE patients. Extrapolating these findings to those in extremis (e.g., cardiogenic shock, altered mental status) should be done cautiously.

Discussion

This trial adds meaningful insight into a common real-world dilemma: can high-flow nasal cannula (HFNC) replace NIV in acute cardiogenic pulmonary edema (ACPE)? The authors demonstrated no significant difference in respiratory rate reduction, dyspnea score improvement, blood gas correction, or hospitalization rates at 120 minutes between HFNC and NIV. While not powered for mortality or intubation rates, their findings support HFNC performing similarly to NIV in the early management of ACPE.

PEEP vs High Flow: Theoretically, NIV should outperform HFNC in pulmonary edema because it provides higher positive pressure (PEEP) to recruit alveoli and reduce left ventricular afterload. However, this study suggests that for moderate ACPE, the modest PEEP generated by HFNC (approx. 2–5 cmH₂O depending on mouth closure) combined with high-flow dead space washout is clinically equivalent to NIV in the first two hours.

Why Patients Switched: The pattern of treatment failure tells the real story regarding the 29 crossovers. Patients switched from HFNC to NIV (16) exclusively due to physiological failure (lack of efficacy or clinical impairment). In contrast, most switches from NIV to HFNC (13) were driven by mask intolerance/discomfort (8), rather than clinical failure. This suggests HFNC may be a reasonable starting point for some patients, but clinicians must be ready to escalate given NIV is able to provide more physiological support and has less clinical failure.

Who Still Needs NIV?: HFNC is not a “set and forget” strategy for the sickest patients. It is not ideal for severe presentations such as:

◦ SCAPE or sympathetic crashing pulmonary edema

◦ Severe hypercarbia with pH <7.25

◦ Impending respiratory fatigue.

In these cases, NIV remains the preferred strategy, as positive pressure is physiologically necessary to recruit alveoli and unload the left ventricle. Furthermore, patients with altered mental status (GCS ≤13) were specifically excluded from this trial. In this population, NIV carries a high risk of aspiration, and early endotracheal intubation is often the safest course.

Medical Management: A major confounder is the medical management parameters. The study notes patients received “standard” diuretics and nitrates but provides no protocol on dosing. In SCAPE and ACPE, aggressive medical bundles (diuretics, vasodilators and afterload reduction) often do the “heavy lifting” in physiological improvement. If medical management was aggressive, the choice of oxygen device may become secondary; if medical management was weak or delayed, the device matters more. Without these details, it is hard to know if HFNC worked because it is great, or because the drugs worked regardless of the device.

Resources: One of the most interesting points in the trial was the exclusion of two patients because the hospital physically ran out of ventilator sets. This highlights a critical logistical and resource challenge. If an academic center with a resource supported trial can run low on NIV, smaller community sites may struggle with the same in surge scenarios. At institutions like mine (UTHealth Houston), HFNC is readily available in the ED and step-down units — even still, NIV remains the default in severe cardiogenic pulmonary edema, particularly when nitroglycerin/BiPAP bundles are implemented. HFNC importantly requires less intensive nursing/RT supervision than NIV. In resource-limited settings (or pandemics), HFNC can be a viable first step if closely monitored.

Patient Tolerance: From a practical standpoint, HFNC allows patients to speak, eat, and clear secretions—activities impossible with a tight-fitting NIV mask. For elderly patients or those in palliative scenarios, this comfort factor may outweigh the marginal physiologic benefits of NIV.

Short-Term Data Only: The study’s 120-minute cutoff creates a significant blind spot regarding “late failure.” While HFNC was non-inferior within the first two hours, we lack data on downstream markers like ICU admission, length of stay, or mortality. For the patient, this implies that HFNC is effective at stabilizing the acute presentation, but we lack evidence that it provides the longer term stability needed to prevent intubation if the pulmonary edema does not resolve rapidly with medical management. In this way it may act as an early bridge therapy in more prolonged cases.

In Practice: In a high-volume ED, HFNC can offer a logistical advantage by being easier to initiate and better tolerated than NIV. However, this ease of use must not lead to a “set and forget” mentality. In this trial, nearly 18% of HFNC patients required escalation to NIV due to lack of efficacy or clinical deterioration. In my clinical bedside practice, there are routine patients who fail HFNC despite meeting criteria and tolerating it initially. Therefore, while HFNC is a valid first-line option or “bridge” for moderate ACPE, it mandates strict reassessment. If the patient hasn’t shown clear clinical and physiologic improvement, clinicians must be ready to immediately escalate to NIV or intubation. One may argue that given the resources associated with escalation, in a busy low resource environment it may be reasonable to start on NIV and de-escalate to HFNC as patients clinically improve.

Author's Conclusion

“In this study, no difference was found between HFNC and NIV in reducing the symptoms and signs of respiratory failure with oxygen-ventilation support in patients with acute cardiogenic pulmonary edema. Considering that HFNC provides better patient tolerability and comfort, it can be considered as an alternative to NIV in this specific patient population.”

Our Conclusion

HFNC is a safe, effective, and well-tolerated alternative to NIV for the initial management of moderate ACPE. It achieves similar improvement in respiratory rate and dyspnea without the discomfort of a tight-fitting mask. However, clinicians must maintain the ability to rapidly escalate to NIV or invasive ventilation if patients fail to show clear clinical improvement within 60 minutes. HFNC or NIV is not an isolated replacement for concomitant medical therapy.

Clinical Bottom Line

High-Flow Nasal Cannula serves as an effective bridge or alternative to NIV for initial therapy in mild-to-moderate ACPE, balancing patient comfort with physiologic support. However, it is not a “set-and-forget” strategy; clinicians must enforce strict reassessment to ensure rapid escalation to NIV if the patient fails to stabilize.

Guest Authors

Dorian Alexander, MD

Associate Professor of Emergency Medicine, Internal Medicine and Critical Care

UT Health, Houston, TX

Ryan Johnson, MD

PGY 3 Emergency Medicine Resident

UT Health, Houston, TX

Showing Slide 1 of 2

References

Marjanovic N, et al.

High-flow nasal cannula oxygen versus noninvasive ventilation for the management of acute cardiogenic pulmonary edema: a randomized controlled pilot study.

Eur J Emerg Med. 2024

PMID: 38364020

Altunbas E, et al.

Is high-flow nasal oxygen as effective as non-invasive ventilation in acute cardiogenic pulmonary Edema?

Am J Emerg Med. 2025

PMID: 40848477

Post Peer Reviewed By: Mark Ramzy, DO (X/IG: @MRamzyDO), and Anand Swaminathan, MD (X/IG: @EMSwami)

Your Deep-Dive Starts Here

Rethinking Fluid Resuscitation in Vaso-Occlusive Crisis: Is Lactated Ringer’s the Superior Choice?

Background: Sickle cell disease (SCD) affects an estimated 300,000 infants born worldwide each ...

Hematology and Oncology

Read More

REBEL Core Cast 126.0 – Peds Hem Onc Emergencies

Take Home Points Early administration of antibiotics (within 60 min) in patients with ...

Hematology and Oncology

Read More

Hematologic Emergencies in the Critical Care Setting Part 2

Background: Hematologic emergencies in the critical care setting are rare but deadly complications ...

Hematology and Oncology

Read More

Hematologic Emergencies in the Critical Care Setting Part 1

Background: Hematologic emergencies in the critical care setting are rare but deadly complications ...

Hematology and Oncology

Read More

REBEL Core Cast 41.0 – Acute Chest Syndrome

Take Home Points 100k people in US have sickle cell, the majority will ...

Cardiovascular

Read More

Showing Slide 1 of 5: Rethinking Fluid Resuscitation in Vaso-Occlusive Crisis: Is Lactated Ringer’s the Superior Choice?

The post HFNC vs. NIV in Acute Cardiogenic Pulmonary Edema: Go with the Flow? appeared first on REBEL EM - Emergency Medicine Blog.