High-flow nasal cannula (HFNC) does generate positive end-expiratory pressure, but the amount is modest and variable compared to what machines like CPAP or mechanical ventilators deliver. In a healthy adult breathing with their mouth closed, each 10 liters per minute of flow adds roughly 0.7 cm H2O of pressure, topping out around 3 cm H2O at typical clinical flow rates. Open your mouth, though, and that number gets cut roughly in half. This “PEEP effect” is real enough to improve oxygenation and lighten the work of breathing in many patients, but it is only one piece of what makes HFNC useful, and the pressures it produces depend on factors that clinicians cannot always control.
How HFNC Creates Airway Pressure
HFNC is fundamentally an open system. Unlike CPAP, which seals around the nose or face and holds a set pressure, HFNC delivers heated, humidified gas through short prongs that sit loosely inside the nostrils. There is no seal. Yet the sheer volume of gas flowing through the nasal passages resists the patient’s exhaled air, and that resistance translates into a measurable rise in pressure at the end of expiration. Researchers have consistently described this as a “PEEP effect” rather than true PEEP, because the pressure is not actively regulated or guaranteed to stay at a fixed level.1Acute and Critical Care. High-flow nasal cannula for respiratory failure in adult patients – Section: Positive End Expiratory Pressure Effect
The relationship between flow rate and pressure is not a simple straight line. Modeling work has found that PEEP rises roughly in proportion to the square of the flow rate, meaning you can double the pressure by increasing flow by about 40%.2PubMed. Upper airway pressure distribution during nasal high-flow therapy In practice, all of the downstream effects of HFNC, including the pressure it generates, CO2 washout, and the fraction of oxygen actually reaching the lungs, are flow-dependent and maximized when the delivered flow exceeds the patient’s own peak inspiratory flow.3PubMed Central. The effects of flow settings during high-flow nasal cannula support for adult subjects: a systematic review That peak flow varies widely between patients and disease states, which is one reason the PEEP effect is hard to predict from flow settings alone.
Mouth Open Versus Mouth Closed
If there is one variable that dominates the PEEP conversation, it is whether the patient’s mouth is open or closed. With the mouth shut, the nasal passages form a relatively closed circuit and the pressure builds. With the mouth open, exhaled gas escapes freely and airway pressure drops substantially. In adults, a closed mouth yields about 0.7 cm H2O of PEEP per 10 liters of flow, while an open mouth yields roughly half that, around 0.35 cm H2O per 10 liters.1Acute and Critical Care. High-flow nasal cannula for respiratory failure in adult patients – Section: Positive End Expiratory Pressure Effect
A recent computational study made the gap even more dramatic in patients with acute hypoxemic respiratory failure. When models were matched to individual patient data from a group of 58 adults receiving HFNC at 60 liters per minute, the average PEEP was about 8.9 cm H2O with mouths closed but only about 1.4 cm H2O with mouths open.4Respiratory Research. Airway pressures generated by high flow nasal cannula in patients with acute hypoxemic respiratory failure: a computational study That is an enormous difference, and it means the clinical reality at the bedside hinges partly on something no machine can enforce: whether the patient keeps their mouth closed. Clinicians sometimes coach patients to breathe through the nose, but in respiratory distress, mouth breathing tends to win.
Interestingly, the open-mouth position is not entirely a loss. While it tanks the PEEP effect, it appears to optimize dead-space CO2 washout, because the continuous flow of fresh gas flushes stale air out of the nasal passages and upper airway more efficiently when air can exit through the mouth.5International Journal of Critical Care and Emergency Medicine. Things to Keep in Mind in High Flow Therapy: As Usual the Devil is in the Detail So the two main benefits of HFNC, pressure and CO2 clearance, are somewhat at odds depending on mouth position, and neither clinicians nor patients can always have both at once.
How Lung Disease Changes the Pressure Picture
The pressures quoted in most textbooks and review articles, usually a few centimeters of water, come from studies of relatively healthy airways or mild respiratory illness. In sicker lungs, the story changes. As functional lung volume shrinks from consolidation or collapse (as it does in severe pneumonia or ARDS), the airway pressure generated by the same HFNC flow rate goes up. The computational study mentioned above found that when half of the model’s lung compartments were non-aerated, PEEP at 60 liters per minute with the mouth closed reached about 11.4 cm H2O, well into the range clinicians typically associate with mechanical ventilation or CPAP.4Respiratory Research. Airway pressures generated by high flow nasal cannula in patients with acute hypoxemic respiratory failure: a computational study
This finding is a double-edged sword. On one hand, it suggests HFNC may be more physiologically useful than its modest bench-test numbers imply for the patients who need it most. On the other, it means the airway pressure is genuinely unpredictable. Two patients on identical flow settings could be experiencing very different pressures depending on how much lung is actually participating in gas exchange. Since there is no way to directly measure the pressure being generated without invasive monitoring, clinicians are essentially flying somewhat blind on this variable.
Dead Space Washout and Breathing Effort
The PEEP effect gets most of the attention, but it may not even be the most clinically important thing HFNC does. A strong argument can be made for dead space washout, the flushing of CO2-rich air from the nose, pharynx, and upper trachea so that each breath the patient takes is richer in oxygen and lower in CO2. In a case report involving a patient with COPD, nasal high flow instantly cut the volume of CO2 being rebreathed, reduced the patient’s breathing rate from 26 to 22 breaths per minute, and lowered their blood CO2, even though PEEP only increased by about 1 cm H2O and tidal volumes stayed the same.6PubMed Central. Nasal high flow reduces hypercapnia by clearance of anatomical dead space in a COPD patient The improvement in CO2 had to come from better dead space clearance, because the patient was actually breathing less total air per minute.
Lab models using 3D-printed airways have confirmed this mechanism. With an open mouth, even a relatively low HFNC flow of 10 liters per minute was enough to dramatically reduce CO2 concentrations below the vocal cords in a normal-lung model. With the mouth closed, much higher flows (40 to 60 liters per minute) were needed to achieve similar CO2 clearance.7PubMed Central. A high-flow nasal cannula system with relatively low flow effectively washes out CO(2) from the anatomical dead space in a sophisticated respiratory model made by a 3D printer This is the flip side of the mouth-position trade-off: open mouth is bad for pressure but good for CO2 washout.
The net result is that HFNC consistently reduces how hard patients have to work to breathe. In a study of patients with COPD who had elevated CO2, the esophageal pressure swings that reflect breathing effort were lowest at 30 liters per minute and comparable to those on non-invasive ventilation. Interestingly, cranking the flow up to 50 liters per minute actually increased effort in half the patients, a reminder that more flow is not always better.8Annals of Intensive Care. Effects of high-flow nasal cannula and non-invasive ventilation on inspiratory effort in hypercapnic patients with chronic obstructive pulmonary disease: a preliminary study
What Matters in Neonates and Small Infants
HFNC is widely used in neonatal and pediatric care, but the physics play out differently in tiny airways. Flow rates are scaled by body weight, and the relationship between the size of the nasal prongs and the size of the nostrils (the prong-to-nares ratio) becomes a critical determinant of how much pressure builds up. In preterm infants, researchers found that body weight and cannula type interact in unexpected ways: at flow rates of 2 liters per kilogram per minute or higher, pharyngeal pressure correlated positively with body weight when using certain cannula designs but negatively with others.9PubMed. Pharyngeal pressure due to high-flow nasal cannula devices in preterm infants Reviews of the pediatric literature confirm that PEEP is influenced by the HFNC device brand, the prong-to-nares ratio, mouth status, and even the type of gas being delivered.10European Respiratory Review. The effects of flow settings during high-flow nasal cannula oxygen therapy for neonates and young children
In pediatric critical care, a cost-effectiveness analysis comparing HFNC with CPAP found that HFNC was cheaper, saving several thousand pounds per patient over six months, with no meaningful difference in quality-of-life outcomes.11PubMed Central. Cost-effectiveness of high flow nasal cannula therapy versus continuous positive airway pressure for non-invasive respiratory support in paediatric critical care The clinical outcomes being similar is the important part: if HFNC delivers enough pressure and respiratory support to match CPAP in many pediatric scenarios, and costs less, the case for using it as a first-line approach becomes practical as well as physiological.
After the Breathing Tube Comes Out
One of the strongest clinical applications for HFNC sits outside the question of whether it provides “enough” PEEP and focuses instead on what happens after a patient is extubated from a ventilator. The transition from mechanical ventilation to unassisted breathing is a vulnerable window. A randomized trial of over 500 low-risk patients found that those given HFNC after extubation were reintubated within 72 hours about 5% of the time, compared to roughly 12% in the group receiving standard oxygen therapy.12PubMed. Effect of Postextubation High-Flow Nasal Cannula vs Conventional Oxygen Therapy on Reintubation in Low-Risk Patients: A Randomized Clinical Trial Post-extubation respiratory failure was also less common with HFNC.
During the COVID-19 pandemic, this application became even more relevant. In a study of COVID-19 patients with acute respiratory failure who had been on mechanical ventilation, post-extubation HFNC was independently associated with lower 28-day mortality after planned extubation.13PubMed Central. High-Flow Nasal Cannula Application After Extubation in Acute Respiratory Failure Patients Whether the benefit comes primarily from the PEEP effect, from dead space washout, from better humidification keeping airways clear, or from patients simply tolerating HFNC better than a face mask (and therefore keeping it on) remains an open question. HFNC systems are generally better tolerated than traditional high-flow oxygen setups like non-rebreather masks, which helps with adherence.14CHEST. High-Flow Nasal Cannula: Mechanisms of Action and Clinical Applications
Knowing When HFNC Is Not Enough
Because HFNC delivers an unpredictable and generally modest level of PEEP, clinicians need a way to recognize when a patient is failing on it and needs escalation to non-invasive ventilation or intubation. Delaying that escalation can be dangerous. The most widely studied tool for this is the ROX index, which combines a patient’s oxygen saturation-to-FiO2 ratio with their respiratory rate into a single number that can be tracked over time.
A meta-analysis of the ROX index in COVID-19 patients found that it performed reasonably well at predicting HFNC failure, with an overall diagnostic accuracy that was useful but not perfect. The optimal cutoff hovered around 5.2, and the index performed differently depending on when it was measured: a lower score in the first 6 to 12 hours was more useful for flagging patients likely to fail, while a higher score at 12 to 24 hours was better for confidently ruling out failure.15PubMed Central. Performance of the ROX index in predicting high flow nasal cannula failure in COVID-19 patients: a systematic review and meta-analysis In infants with bronchiolitis, the ROX index also showed discriminating ability from about 12 hours onward, with failure and success groups showing clearly different scores.16Scientific Reports. ROX index as a predictor of failure of high-flow nasal cannula in infants with bronchiolitis
The broader takeaway from this research is that HFNC should not be treated as a set-and-forget therapy. The ROX index and similar monitoring tools exist precisely because the level of respiratory support HFNC provides, including its PEEP effect, may be inadequate for a given patient’s needs, and the only way to know is serial reassessment.17PubMed Central. Prediction of high-flow nasal cannula failure in critically ill patients: a narrative review
Infection Control and Aerosol Dispersion
A practical concern that emerged forcefully during the COVID-19 pandemic is whether HFNC sprays infectious aerosols farther than standard oxygen therapy. The answer, based on experimental work, is that exhaled air velocity during HFNC is higher than during standard oxygen delivery, and the estimated dispersion distance for large droplets is correspondingly greater.18European Respiratory Journal. High-flow nasal cannula for COVID-19 patients: risk of bio-aerosol dispersion This led to early reluctance to use HFNC in COVID-19 wards for fear of exposing healthcare workers.
However, computational fluid dynamics simulations later showed that placing a simple surgical mask over the patient’s nose and mouth during HFNC captured nearly 89% of exhaled particle mass at 40 liters per minute. That was actually more than the roughly 77% captured during low-flow oxygen with a mask and more than the 73% captured during normal tidal breathing with a mask.19JACEP Open. Reducing aerosol dispersion by high flow therapy in COVID‐19: High resolution computational fluid dynamics simulations of particle behavior during high velocity nasal insufflation with a simple surgical mask The high flow essentially pushes exhaled particles outward into the surgical mask, where they deposit. Without any mask, the picture is worse: during unmasked normal breathing without therapy, over half of exhaled particles travel beyond one meter. The practical message is that HFNC with a surgical mask worn over it is a reasonable infection-control strategy, arguably better than some alternatives.
Delivering Nebulized Medications Through HFNC
An increasingly explored use of the HFNC circuit is as a vehicle for delivering aerosolized medications, particularly bronchodilators, without interrupting oxygen therapy. The appeal is obvious: a patient who needs both continuous high-flow oxygen and periodic inhaled medications does not have to switch devices. The technical reality, though, is more complicated than plugging in a nebulizer.
Studies have converged on a few practical points. Vibrating mesh nebulizers consistently outperform older small-volume jet nebulizers when plumbed into an HFNC circuit.20PubMed. The Impact of High-Flow Nasal Cannula Device, Nebulizer Type, and Placement on Trans-Nasal Aerosol Drug Delivery Placement matters too: positioning the nebulizer upstream near the humidifier chamber generally delivers more drug to the lungs than placing it close to the nasal cannula. And the flow rate itself creates a paradox. At lower flows, a higher fraction of the nebulized drug reaches the patient’s lungs during calm breathing. At higher flows, the turbulence scatters more of the aerosol and less makes it through. The sweet spot for drug delivery appears to be when the HFNC flow is set below the patient’s inspiratory flow, with diminishing returns as gas flow approaches about half of what the patient is pulling in per breath.21PubMed Central. Aerosol delivery through high-flow nasal therapy: Technical issues and clinical benefits
Newer approaches like inspiration-synchronized nebulizers, which pulse drug delivery to match the patient’s inhalation, have shown promise in bench testing. One study found that synchronizing a small-particle vibrating mesh nebulizer with inspiration consistently improved drug delivery compared to running the nebulizer continuously, particularly at moderate flow rates.22PubMed. Aerosol Delivery to an Adult Model via High-Flow Nasal Cannula with Inspiration-Synchronized Small-Particle Vibrating Mesh Nebulizer Even with these refinements, the total percentage of loaded drug that reaches the lungs via HFNC is small. But the bronchodilator effect of drugs like albuterol appears to be clinically similar to what you get from conventional delivery methods, suggesting the drug that does arrive is getting where it needs to go.