Mask Selection in CPAP and NIV: Mask On. Problem Solved?

Anyone who has been fitting masks for years develops a remarkably good instinct. You see a face, know the models, and often already have a suitable solution in mind. And yes: often you are right.

That is experience. And it is valuable.

But honestly: how often do we reach for a model because we know it particularly well? Because it is available? Because it worked for the last ten patients?

Of course, nothing speaks against a tried-and-tested model as a starting point. The assumption that it must therefore also fit the eleventh patient is another matter.

When does experience quietly turn into habit?

The question is not whether we should unpack the entire mask range every time. On the contrary: good preselection saves time and unnecessary trials. What matters is that we know why we are suggesting this particular mask, and that we remain willing to revise our assessment.

Especially at the start of therapy, first impressions count. A patient who gets air blown into their eyes at night, wakes up with pressure marks, or feels constricted by the mask does not exactly start out enthusiastic. Equally important is safe handling: can the patient put the mask on independently and remove it quickly and without difficulty if needed? Especially during the acclimatisation phase, this sense of security can reduce fear and uncertainty.

The first experiences with the mask in particular can shape acceptance of the therapy going forward.

Whether the patient wears a nasal mask, nasal pillows, or an oronasal mask is not just a matter of personal preference. It can also affect the effectiveness of the therapy.

In CPAP therapy in particular, a closer look is worthwhile. A large analysis with a total of 6,378 participants found that nasal masks and nasal pillows often performed more favourably than oronasal masks: residual AHI was lower, less therapy pressure was needed, and average usage duration was also higher [2].

Does this mean the oronasal mask is fundamentally the worse choice? Of course not.

The strength of the evidence is limited, and what works better on average does not automatically apply to every patient.

For non-invasive ventilation (NIV) in long-term care, the evidence is less clear-cut. An analysis of 290 patients with COPD or obesity hypoventilation syndrome found no significant differences between nasal and oronasal masks regarding blood gas values or daily usage duration [3].

And mouth breathing? It, too, is not automatically an argument for an oronasal mask. What matters is whether relevant mouth leak is actually present and how well nasal breathing functions [1, 4].

So the decisive question is not which mask type is fundamentally the best, but which one works best for this patient and their therapy goal.

Awake, upright, head straight. The mask seals, the leak is unremarkable. All good. For now, at least.

But what happens once the patient lies down? The change in position alone can alter how the mask sits on the face. During sleep, the muscles also relax, which can let the jaw drop and the mouth open. If the patient turns onto their side or stomach, the pillow may additionally press against the mask. All of this can promote leaks.

What seals perfectly sitting up may not work at all lying down.

For this reason, mask fit should ideally also be checked lying down, under therapy pressure, and in the patient's usual sleep position [4].

And dentures? Are they worn at night or removed? Especially with oronasal masks, this can change the facial contour and compromise the seal.

Beard growth, scars, facial asymmetries, or heavy sweating can also make sealing more difficult.

But not every mask problem shows up as escaping air. Sometimes the mask leaves quite different traces.

The bridge of the nose? We usually spot that right away. But what about the back of the neck, the back of the head, or the areas behind the ears?

Mask straps, too, can cause pressure and friction damage. Especially under longer hair, early skin changes can easily go unnoticed. Redness or pressure discomfort alone should already prompt a closer look, not only once the skin is broken.

For pressure marks on the bridge of the nose, oronasal masks with under-the-nose cushions can be a sensible alternative.

But not every pressure problem calls for a new model right away. A mask should seal as well as possible, but not at any cost. Often it already helps to correct the mask fit or adjust strap tension to relieve individual pressure points. Suitable protective pads can additionally help, provided they do not create new pressure points or compromise the seal [4,8].

A mask does not become tighter simply because we keep pulling the straps harder.

We all have favourite models. That is only human, especially when we get good results with them. Suppliers, too, must take their range, availability, and operational logistics into account. That is legitimate and says nothing about the quality of their work.

But if the first choice does not work, the search must not end with the familiar model or manufacturer. The best option for the patient should take precedence over our personal preference.

And the patient themself? They have an opinion too.

Some want as little material on the face as possible; others feel more secure with a larger mask. The handling of the clips can also be decisive in the choice.

Limited manual dexterity, reduced sensation in the hands, or lacking hand strength can turn even simple clips into a challenge. Here, a different clip mechanism alone can make the difference.

Safety matters too. Magnetic clips can make handling easier but are contraindicated with certain implants or medical devices, such as pacemakers. Non-magnetic variants are now available for various mask models. The manufacturer's own safety instructions remain decisive here, including with regard to people in close bodily contact [5].

A mask can fit technically excellently and still be the wrong one for this patient.

Before opening the next mask packaging, it is worth questioning the complaint more closely. Is it really the mask that is felt to be uncomfortable? Or could it be the therapy settings?

In CPAP, for example, an uncomfortable sense of pressure or unsuitable ramp settings can play a role. In NIV, ventilation settings not sufficiently adapted to the patient can likewise make the therapy harder to tolerate [7].

For this reason, therapy data and device settings should be reviewed professionally alongside mask fit. Sometimes we change the mask even though the actual cause lies elsewhere entirely.

And what if, despite careful selection and fitting, no suitable off-the-shelf model can be found? A custom-made mask can then be a good option. But this kind of provision comes with considerable effort and high cost.

Custom-made, however, does not automatically mean comfortable. Rigid, moulded models can feel heavy or hard. If the facial contour changes over time, renewed reworking may become necessary, and that means additional effort.

A recent randomised 2026 study of 160 adults with obstructive sleep apnoea examined individually made, silicone-based oronasal mask cushions. It showed no superiority over conventional interfaces. In the study group, leaks were in fact higher and adherence lower. This cannot be generalised to all moulded custom masks, but it does underline that custom fabrication does not automatically mean better therapy outcomes [6].

How far such provision can go is illustrated by an unusual case from practice:

A patient had already tried numerous mask models without finding a truly suitable solution. Sometimes it was leaks, sometimes an uncomfortable sense of pressure or other problems that prevented satisfactory care. In the end, the decision was made for a custom-made mask.

But even then, the difficulties were not over. Even the mask made specifically for her caused uncomfortable pressure sensations. It was reworked, corrected, and refitted again. Repeatedly. When one spot was relieved, another started pressing. So, rework again. Refit again. And hope again that this time it would work.

But it did not stop at mask fitting alone. The therapy settings were also reviewed and adjusted several times over the course of treatment, to make the therapy as tolerable as possible for the patient.

The care process dragged on for nearly a year. With great effort from everyone involved, considerable cost, and substantial personnel time. And yet the therapy was barely used during this period.

Eventually, a different question arose: What if the real problem was neither the mask nor the device settings?

We increasingly had the impression that the difficulties might lie deeper. In an open, empathetic conversation, we therefore stopped talking only about masks and settings and started talking about how the patient herself experienced the therapy. In doing so, she became increasingly aware that she was not merely struggling with the mask, but that she actually did not want to continue the treatment at all.

Her complaints were real and had to be taken seriously. But technical corrections alone could not resolve this underlying conflict. What was needed here was something else: understanding, honest information, and the willingness to respect an informed decision against the therapy as well.

Sometimes it is not the next adjustment that matters, but the next honest question.

This case is an extreme example and certainly not typical everyday care. But it reminds us of something essential: not every case of low therapy usage is a mask problem. And not every mask problem should be hastily dismissed as lack of motivation.

Experience matters. It helps us select masks purposefully instead of letting patients work through half the range. But it must not stop us from repeatedly questioning our own decisions.

Perhaps a patient really does need the sixth mask. Or perhaps the second one would already have worked, had we looked more closely. Perhaps the problem is not the interface at all, but the therapy settings. Or we have not yet understood well enough how the patient themself feels about their treatment.

We will not find out by simply opening the next package.

Sometimes the best decision in mask selection is not to select a new mask yet.

Can digital facial scanning meaningfully support mask preselection? And where are its limits? We take this up in the next article.