NIV Interfaces, Helmets and Mask Fit

In depth

The one-page summary above is what the PDF prints. These sections go further.

Interfaces compared

An interface is chosen for the patient's breathing route, face, alertness and expected hours of use. In acute respiratory failure most patients breathe through the mouth, so BTS/ICS 2016 makes an oronasal mask the usual first interface. It controls mouth leak, at the cost of more claustrophobia, difficulty speaking, eating and clearing secretions, pressure on the bridge of the nose, and a risk of aspiration if the patient vomits into it. Mask names vary: UK guidelines call the oronasal mask a 'full face mask', and some US papers use 'full-face' for the total face mask.

A nasal mask or nasal pillows leave the mouth free, so the patient can talk, cough and drink, and nasal masks have less dead space. BTS/ICS advises that a patient who finds the oronasal mask claustrophobic or distressing can be tried on a nasal mask or nasal pillows by an experienced team. They work only while the mouth stays closed: mouth leak limits them during sleep, a chin strap holds the jaw but cannot seal the lips, and they do not suit a mouth breather or a blocked nose. If mouth leak prevents effective ventilation, return to an oronasal mask.

A total face mask seals around the perimeter of the face (eyes included, ears not), so it puts no direct pressure on the bridge of the nose. BTS/ICS finds it useful when leak stays excessive or nasal bridge ulceration develops. In a before-and-after study in 5 ICUs (Schallom 2015, 200 patients), pressure ulcers developed in 20% of patients with an oronasal mask and 2% with a total face mask. It still covers the mouth, so the vomiting risk is unchanged and close monitoring is needed.

A mask over the mouth can trap vomit. Masks that cover the mouth are therefore avoided in a patient who cannot remove the mask unaided, for example with arm weakness from neuromuscular disease (Praud 2025). A nasal mask, with a chin strap if the mouth opens, is the safer choice when the mouth stays closed in sleep. The Swiss home ventilation recommendations add that when an oronasal mask is the only way to control mouth leak, the patient should be given a means of pulling it off, such as a hand strap.

Fit, leak and the skin

The first minutes decide whether the patient accepts the mask. Choose the size with the manufacturer's sizing gauge, and let the patient hold the mask against the face before the straps are fastened (Merck). Typical starting pressures are IPAP 10–12 and EPAP 5–7 cm H₂O with the patient upright or semirecumbent; reassess from 30 minutes and raise IPAP as needed. Coaching comes before sedation: NIV depends on an alert, cooperative patient, and sedation can hide the signs of failure.

Tighten the straps evenly and moderately, then make the final adjustment with the patient lying down and the pressure on. Merck allows one or two finger widths under the strap. Masks are built to mould to the face when pressurized, and over-tightening stops them doing so; BTS/ICS names over-tightening as a common cause of nasal bridge ulceration. Pressure measured on the nasal bridge is lower supine than seated (Brill 2018), which is one reason to finish the fit in the position the patient will sleep in.

A small leak, such as 5 L/min, is negligible, so early redness on the nasal bridge is a reason to protect the skin, not to tighten the straps. Most ventilators compensate for leak; a large leak is different, because it drops the delivered pressure, causes auto-triggering and delayed cycling, and needs the mask repositioned or changed.

The bridge of the nose has little tissue over bone and is the usual site of injury; BTS/ICS reports nasal bridge ulceration in 5–10% of patients, and it can end NIV. Look at the skin under the mask every 2–4 hours. At the first sign of skin trauma, BTS/ICS advises a barrier dressing, regular breaks and alternating between two interface types so the pressure points move. If the patient still needs NIV most of the day, a total face mask takes the load off the bridge; in a study of do-not-intubate patients whose oronasal mask NIV was failing, those switched to a total face mask within the first 12 hours developed pressure sores in 24%, against 87% of those switched later (Lemyze 2013).

Circuits, exhalation ports and rebreathing

Critical care ventilators use a double-limb circuit: exhaled gas returns through the expiratory limb, so the mask is non-vented (no exhalation holes) and no minimum PEEP is needed to prevent rebreathing. Bilevel devices use a single limb. Exhaled gas then has to leave through an intentional leak, either holes built into the mask (a vented mask) or a port placed in the circuit close to a non-vented mask, and a minimum EPAP of about 3–4 cm H₂O is needed to flush it out.

Whatever is not flushed out before the next breath is rebreathed. Rebreathing is less with the exhalation port in the mask than with a port in the circuit (Schettino 2003, lung model). It rises with low EPAP, a short expiratory time (tachypnea), a large tidal volume and extra dead space between the mask and the port. When a hypercapnic patient on a single-limb circuit is not improving, check the mask type and the port position along with the settings.

Interface volume is not the same as dead space. In a computational study (Fodil 2011), effective dead space differed only modestly between interfaces (about 110–370 mL) while their internal volumes ranged from about 110 mL to 10 L: a small oronasal mask adds close to its own volume, and the largest interfaces add far less than theirs. Size still counts for something: in Schettino's lung model a total face mask (875 mL) allowed more rebreathing than a 165 mL mask with the port in the mask. Choose by comfort, leak and synchrony first.

Oronasal and total face masks have a built-in anti-asphyxia valve. It is held shut by the circuit pressure and opens to room air when the pressure is lost (below about 3 cm H₂O on one total face mask), so the patient can still breathe if the blower or the power fails. Never block it.

The stomach, vomiting and who should not have a mask

NIV leaves the airway unprotected, so the patient needs enough alertness and airway reflexes to protect it. Merck lists as absolute contraindications: cardiac or respiratory arrest; severe upper gastrointestinal bleeding; facial deformity or trauma; upper airway obstruction; copious secretions or inability to clear them; vomiting or impaired gastric emptying (ileus, bowel obstruction); an imminent need for surgery or a setting where close monitoring is not possible; and obtundation or inability to cooperate. A face that cannot carry a sealed mask (midface fractures, facial burns, severe deformity) rules mask NIV out however good the indication. BTS/ICS treats some of the others as less absolute: confusion or agitation makes NIV harder but should not by itself preclude a trial.

Merck's indications include COPD exacerbation with hypercapnia, cardiogenic pulmonary edema, hypoxemic respiratory failure, obesity hypoventilation, and patients with a do-not-intubate directive who would otherwise need intubation.

Inspiratory pressure above the esophageal opening pressure, about 20 cm H₂O, pushes gas into the stomach, so Merck keeps IPAP below 20 cm H₂O in acute NIV. Belching, nausea and a distended abdomen on NIV mean the stomach is filling: lower the IPAP below 20 cm H₂O, then repeat the blood gas to see whether the lower pressure still controls the PaCO₂. BTS/ICS adds that severe distention may need a nasogastric tube. (Home high-intensity NIV for chronic COPD often runs above 20 cm H₂O; the ceiling here is the acute-care rule.)

If a patient on mask NIV retches or vomits, take the mask off at once and keep the patient upright. BTS/ICS makes rapid removal of the mask the key safety issue with vomiting, which is why the patient must be able to signal or to pull the mask off. Changing the pressure or the interface comes later, once the airway is safe and the team has decided whether NIV can continue; a patient who keeps vomiting or becomes less alert is no longer a candidate.

Helmet NIV

A helmet is a transparent, soft hood that encloses the whole head and seals at the neck with a soft collar; on many models, braces under the arms hold it down when it is pressurized. Nothing presses on the face, so it avoids facial pressure injury, and patients tolerate high PEEP for long sessions. Its main use is moderate to severe acute hypoxemic respiratory failure (PaO₂/FiO₂ under 200) in the ICU; guidelines make no firm recommendation for or against it.

The hood is large and compliant, and that changes how the ventilator behaves. Part of each breath stretches the hood before pressure reaches the airway, so triggering and cycling are delayed and pressure rises slowly. In a lung model, trigger delays were more than twice as long with a helmet as with a face mask and shortened as CPAP or pressure support rose (Moerer 2006). In a crossover study after extubation, a helmet at mask settings increased inspiratory muscle effort, and raising pressure support and PEEP by 50% with the fastest pressurization rate abolished the difference (Vargas 2009). Helmet settings are therefore not mask settings: PEEP about 10–14 cm H₂O, pressure support about 8–14 cm H₂O or higher for strong efforts, and the shortest rise time the ventilator allows.

The tidal volume on the screen includes the gas that distends the hood, and the displayed minute ventilation reflects washout flow, so neither measures the patient's breath. A large displayed tidal volume is not a sign of failure with a helmet, as it can be with a mask. Judge the helmet by gas exchange, respiratory effort and dyspnea, and escalate to intubation promptly if they are not improving. Some CO₂ rebreathing persists even with good settings, so check arterial blood gases.

• Circuit: a double-limb circuit with the inspiratory and expiratory limbs on opposite ports, not a Y-piece on one port, so that fresh gas sweeps through the hood. Raoof 2020 advises an ICU ventilator with dual circuitry and an NIV option, and a filter on the expiratory limb.
• Humidification: usually none for helmet NIV on a ventilator. Breathing inside the closed hood conditions the gas, and BTS/ICS notes that humidified gas rains out in the helmet.
• Leak at the neck: reposition the helmet or change the hood or collar size. Tightening the armpit braces is not the answer: pressure in the axilla can cause vascular complications (axillary vein thrombosis has been reported). Pull the braces downward and secure them to the bed instead.
• Noise and ear pressure: offer earplugs, which ease both.
• Sedation: if light sedation is needed for tolerance, dexmedetomidine is preferred because it changes respiratory drive little; any sedation can mask the signs of failure.
• PEEP: lowering it to quieten the helmet or ease a leak is the wrong fix, because PEEP of about 10 cm H₂O or more is needed to pressurize the hood as well as to treat the patient.

Helmet CPAP and continuous-flow CPAP

In helmet CPAP the fresh gas flow does two jobs: it holds the pressure and it washes exhaled CO₂ out of the hood. The helmet behaves like a semi-closed space in which inspired CO₂ depends on the patient's CO₂ production and the flow through the hood (Taccone 2004). A ventilator in CPAP mode does not supply enough flow, and inspired CO₂ was high when helmet CPAP was given that way; a continuous high-flow source (a flow generator or Venturi system) kept it low. Use at least 50 L/min: in volunteers, inspired CO₂ fell as the flow was raised from 30 to 45 to 60 L/min (Patroniti 2003). A 2021 technical review reads the same studies as showing significant rebreathing below 40 L/min and puts the working flow for helmet CPAP in the range of 60 L/min (Coppadoro 2021). A rising respiratory rate and PaCO₂ in a well-sealed helmet running at low flow point to rebreathing, and the fix is to raise the fresh gas flow. A Venturi flow generator is much louder than the ICU around it and delivers dry gas, so earplugs and a heated humidifier are considered for helmet CPAP.

The same rule applies to continuous-flow mask CPAP used for atelectasis. AARC describes it as breathing from a pressurized circuit against a threshold resistor that holds 5–20 cm H₂O through inspiration and expiration. It needs a gas flow during inspiration sufficient to keep that pressure, from a flowmeter or blender with a reservoir bag on the inspiratory line, or from a demand valve. If the manometer sags on every inspiration and the reservoir bag collapses, the patient's inspiratory flow is outrunning the supply: increase the flow. A different resistor does not help. PEP and EPAP need no gas source, but they apply pressure during exhalation only.

Mouthpiece ventilation

Mouthpiece ventilation is open-circuit daytime support for patients with neuromuscular disease or high spinal cord injury who need a ventilator for part or all of the day. The mouthpiece is held near the lips on a flexible arm fixed to the wheelchair; the patient takes a breath when needed and lets go to speak, eat, swallow or cough. A nasal or oronasal mask is used for sleep. It needs a cooperative patient with enough lip and bulbar function to grab and seal the mouthpiece, and some neck movement, so bulbar weakness is its main limit: in ALS only about 19% of patients can use it successfully.

The 2020 ENMC workshop makes a volume mode the first choice. A set volume lets the patient stack breaths (hold one breath with the glottis closed and take another on top) to reach a deep inflation that strengthens the cough and keeps the chest wall mobile, and it avoids the untimely leak compensation a pressure mode produces each time the mouth leaves the circuit. Pressure modes (12–20 cm H₂O) are a second choice, sometimes preferred for children, and do not allow breath stacking. CPAP is not ventilation and does not assist weak respiratory muscles.

A dedicated mouthpiece mode sets PEEP to 0 and the backup rate to 0, uses a highly sensitive flow-interruption trigger, and turns the disconnection and apnea alarms off, because an open circuit would otherwise alarm constantly; very dependent patients keep an apnea alarm. Suggested starting tidal volumes are 500–600 mL for an adult with Duchenne muscular dystrophy and about 1,000 mL in ALS; if a breath feels too large the patient lets the excess leak out of the mouth.

Home NIV: humidity, rainout, a blocked nose and CPAP that is not enough

Most home NIV users have a heated humidifier; it limits mucosal drying and the rise in nasal resistance that mouth leak causes. Match the setting to the complaint. A dry nose and mouth: turn the humidifier up. Rainout, water in the mask and gurgling in the tubing, happens when humidified gas cools in the tube: turn the humidifier down, and consider heated tubing, an insulating tube cover, a warmer room or placing the device lower. Condensate can also disturb triggering and pressurization. A hose holder that takes the weight of the tubing off the mask reduces leak and skin injury and keeps water from running into the mask.

Mouth leak on a nasal mask has two different fixes. If the nose is clear and the mouth simply falls open in sleep, add a chin strap. If the nose is blocked by a cold, a nasal mask no longer suits the patient: change to an oronasal mask until the congestion clears, as long as the patient can take the mask off unaided. A higher humidifier setting, sleeping with the head raised and a nasal decongestant for no more than 5 days also help.

Obesity hypoventilation: the 2019 ATS guideline suggests CPAP first for stable ambulatory patients who also have severe obstructive sleep apnea (AHI of 30 or more). CPAP holds the upper airway open but does not increase alveolar ventilation. If symptoms have not resolved, or gas exchange awake or asleep has not improved enough on CPAP that is being used well, the guideline's pathway changes to bilevel NIV, titrated with overnight monitoring. Patients with less severe sleep apnea, worse lung function or more severe hypercapnia at the start are the ones less likely to respond to CPAP, and the first 2 months are the time to check.

After surgery, chest trauma and comfort care

Surgery near the diaphragm, anesthesia and pain lower lung volumes and cause atelectasis, and diaphragm dysfunction can last up to 7 days. CPAP and bilevel NIV both restore aeration, and ERS/ATS 2017 suggests NIV for postoperative acute respiratory failure (conditional recommendation, moderate certainty): pooled trials showed lower mortality, fewer intubations and less nosocomial pneumonia. In the NIVAS trial (Jaber 2016, 293 patients with hypoxemic failure after abdominal surgery), face-mask NIV lowered reintubation within 7 days from 45.5% to 33.1% compared with standard oxygen. Surgical complications such as an anastomotic leak or intra-abdominal sepsis are dealt with first, and the patient must be cooperative and able to protect the airway.

The 2022 ERS high-flow guideline sorts postoperative patients by their risk of pulmonary complications: either HFNC or conventional oxygen at low risk, and either HFNC or NIV at high risk (both conditional). After cardiothoracic surgery, HFNC did no worse than intermittent NIV in 830 patients with or at risk of respiratory failure (Stéphan 2015). HFNC is also the noninvasive option left when mask NIV is contraindicated: vomiting, ileus and bowel obstruction rule out a sealed mask (Merck), while a nasal cannula leaves the mouth free. ERS 2022 makes the same allowance after extubation, where it prefers NIV to HFNC for high-risk patients unless NIV is contraindicated. No guideline has tested HFNC in patients with ileus, so watch closely, and intubate promptly if the patient is not improving.

Chest trauma: ERS/ATS 2017 suggests NIV for chest trauma patients with acute respiratory failure (conditional recommendation, moderate certainty). Pooled trials showed fewer intubations, less nosocomial pneumonia and an ICU stay about 2.5 days shorter. The panel advises a cautious trial when pain is controlled and hypoxemia is not severe, so effective analgesia comes first. Positive pressure can cause barotrauma, including tension pneumothorax: BTS/ICS advises that an acute pneumothorax be drained before NIV is applied, and unexplained distress or chest pain on NIV means a pneumothorax must be excluded.

Palliation: ERS/ATS 2017 suggests offering NIV to dyspneic patients for palliation in terminal cancer or other terminal conditions (conditional recommendation, moderate certainty). In pooled trials NIV lowered dyspnea scores and morphine requirements, so it can ease breathing while leaving the patient more awake than higher opioid doses would, which matters to someone who wants to stay lucid to talk with family. The guideline separates the patient who declines intubation but wants NIV as treatment, hoping to survive the admission, from the patient whose only goal is comfort. For comfort, NIV is working if it relieves breathlessness without distress from the mask and without unduly prolonging dying. Patient selection and trained staff matter.