Precision and performance testing of popular CPAP devices [ENG]

The first test I wanted to do when creating was a test of the stability of the pressure achieved by the devices. I even constructed my own measuring equipment in the form of a polycarbonate tube bent in the shape of a U-shaped tube filled with water. However, the improvised measurement method turned out to be too imprecise and my access to the devices was very limited at that time. After less than a year I come back to the subject. This time, however, equipped with professional measuring equipment and a very rich set of CPAP devices.

The test was intended to verify whether the devices we use on a daily basis are precise enough and whether there are noticeable differences between them. Additionally, I decided to test how CPAP deal with leaks. The test limit was deliberately moved to a level where some of the PAP's had problems with the pump performance. It is worth being aware of the fact that this second part of the test was in quite extreme conditions, which most people will not meet in real life.

In case of leakage, the tested parameter is not only the pressure, but above all the flow that the device pump is able to generate. I pay attention to this in each of my tests, although so far it has been more about the comfort of the therapy. As it turns out, this performance is also important at the time of leakage and some market players have decided that the parameters of their devices are too weak. Pay attention to it at the next purchase and put pressure on the manufacturers to improve in the next generations of their machines.

A word of comment on why the parameter of stability and accuracy of pressure is so important for CPAP. Titration does not always take place on the device on which the therapy is carried out. Determining the effective therapeutic pressure at OBS involves finding the smallest pressure at which the respiratory tract expands, allowing free breathing. If the device is unable to maintain this pressure or simply sets it incorrectly, the airway collapses and the therapy is no longer effective. Overpressure should theoretically have less effect. Nevertheless, it is also not recommended, as it increases the side effects of PAP therapy and reduces the comfort of the patient. In particular, awakenings may occur and additional central apnea may appear. Therefore, the stability and accuracy of the pressure in the device is very important for us.

I would like to point out that the vast majority of the CPAP's passed the test without leaks, and this situation should be considered normal. Every manufacturer and physician will always convince you that the right thing to do is not to choose a device that will deal with leaks, but to act to eliminate them, e.g. by proper selection and adjustment of the mask. This is, of course, a very rational! And of course, this does not change the fact that each of us sometimes struggles with incidental leaks. However, they must not be treated as a norm!


We tried our best to guarantee the reliability and objectivity of the test. There were 3 people on the "test committee" and none of them was involved with the CPAP manufacturer. Each result was read by 2 people. As a measuring instrument we used PeakTech 5150, which before the test was checked and calibrated at the manufacturer's service.

All the devices tested by us were new and all the measurements were made in the same conditions. We took all the devices that I had access to at that moment, trying to include all the popular ones. In two cases, higher class devices were included in the test. In the case of ResMed, I did not have access to the popular AirSense and BiPAP AirCurve was tested. However, we confirmed with 100% certainty that from the technical point of view both lines of devices are identical. The changed algorithm should in no way affect the result of our test. The same is true of the Prisma LAB device, which is basically a ventilator - but the pump and measuring system are the same as in the popular Prisma 20a. At a later date we will confirm the results for both of these devices in a lower line

In the case of Prisma, the economic model Prisma Smart was also tested (we wanted to confirm whether there is a difference to the flagship model). Philips put up two competitors in the main competition: the old Remstar and the new Dreamstation. DeVilbiss Blue, whom I'm currently testing, couldn't miss out on the test. China is pushing the market harder and harder and hence in our test two devices: cheap and popular Apex iCH and the highly futuristic Resvent iBreeze Auto. I was very curious whether these devices differ in this field from the more recognized Western manufacturers.

Using the possibility we have additionally tested two more mobile devices: ResMed AirMini and Philips Dreamstation GO. However, their rating will be in a separate category, because it would be unfair to expect the same performance from them as on large, stationary machines.

For the measurement, all of the devices have been put into manual mode. All additional settings such as exhalation relief, ramp, flex and automatic switch-off have been disabled. Before taking the measurement we always waited at least 30 seconds for the device to stabilize (only APEX needed here even 2 minutes).

The first measurement was carried out without any leakage, so in "laboratory" conditions. We decided to measure at 4 pressures:
- 4 cm H20 - the lowest pressure used in therapy
- 10 cm H20 - the most common value in therapy
- 15 cm H2O - maximum value at which CPAP therapy should be performed (BiPAP is already recommended above)
- 20 cm H20 - maximum pressure reached by CPAP's

In the second test we simulated a leakage that occurs when there is a hole in the mask of about 3.5 mm diameter. This generates a significant leakage of 80-120-155-170 l/min depending on the set pressure (4-10-15-20 respectively). These are, of course, quite extreme conditions, but they occur in life and as you will see in the results you can deal with them very well.

We checked the operation with a "normal" leakage of 15-20 l/min, but the results were identical to those of the test without leaks, so I will not present them.


The results of the first test are almost boring. Apart from one, all the devices did very well in this test. Possible deviations are at the limit of the measurement error and are not worth paying attention to. However, Löwenstein's machines, which are clearly the most accurate of all tested devices, deserve a special mention.

The APEX machine was completely compromised here. He was the only one in the whole test who could not set the expected pressure. In addition, it took him most of the time to "get on the speed". I definitely do not recommend this device!


This is where it gets interesting. As you can see, the differences are very clear between the different CPAP's. There are two independent problems here:
- a problem with setting the required pressure (measurement?)
- a problem with the pump performance

That's how it looks in the chart:

Since the above graph is very difficult to analyze, let us look at the distribution of pressures in particular compartments:
At a pressure of 4 cm H2O, all devices perform well... except for the unfortunate APEX.  Prisma beats the competition with precision measurement. Most devices (except Philips Remstar) tend to overpressurize.
Raising the pressure to 10 cm H20 Prisma Smart starts to run out of air. DeVilbiss, Dreamstation and AiCurv- OK, but they tend to overestimate the pressure. Philips for a change lowers the pressure slightly but still achieves a good result.

At 15 cm H20, it is clear that the pumps of most apparatus are no longer running. With the exception of Prisma LAB, which holds the pressure as if enchanted, the DeVilbis and Resmed score (although as you can see overstated) should be indicated as a plus. Philips for a change lowers the pressure in both devices but this is still an acceptable result. Prisma Smart, Resvent and of course APEX don't count anymore
A pressure of 20 cm H2O with a leakage of 170 l/min are conditions that only the more expensive Prisma has managed to cope with. DeVilbis and ResMed are still trying to take up the fight, but they are already running out of air. The rest of the devices are out. 



This device was completely compromised in our test. In my opinion, it should not be registered as a medical device. The results are not set in any way, even without a leak. The differences are so big that the device can be harmful to health!
In addition, in the test we observed the slowest reaching of the target pressure and "waving of indications". It is not a good CPAP and I do not recommend its purchase.

DeVilbiss Blue AutoPlus

Blue is at the top of our list and stands out for its ability to compensate for the highest leaks, which most of the other competitors in our test can no longer cope with. Tendency to overpressurize at medium values, in case of leakage.

Philips- Respironics Remstar Auto

This is the oldest construction in our list. Remstar is sold for several years. What you can see on the chart is very impressive. Our "grandfather" beats most of modern constructions. Up to a value of 15cm/H20 it behaves excellent! Above it begins to lack some air but let's say frankly that this is an extreme situation. Bravo!

Philips Dreamstation Auto

It's not bad, but I felt a little so disappointed, given the high expectations. I expected Philips to make sure that the new generation of their hearing aids would be better than the previous one. Unfortunately, that was not the case. Dreamstation has clear problems with pump performance. In our leak test it was not able to exceed 14 cm H20, and at 10 cm H2O it is not as precise as its older brother.

Löwenstein PRISMA Smart

That's a big disappointment. The low-cost Lowenstein model cannot handle leaks at all and is likely to be a problem with pump performance. If they tell you that this model differs from the more expensive one only by the screen, show them this test. On the plus side for Smart, I must admit that it is mega-precision at the moment when there is no leakage. But as our test showed, it is not art at all...

Löwenstein PRISMA LAB

If you want to accuse me that the test conditions were too demanding, look at the results of this Prisma. If I didn't see it, I wouldn't believe how stable and efficient CPAP can be. In this case, the results are so excellent that you can use Prisma to calibrate your measuring instruments. This is undoubtedly the best CPAP, considering the parameters analyzed in the test. Respect!

ResMed AirCurve VAUTO (AirSense)

Another strong player in our ranking. The course of pressures is similar, although slightly better than in DeVilbiss Blue, with a tendency to slightly overpressurize  when the leakage is at lower values. There is no shortage of power here, but a little more precision would be useful.

Resvent iBreeze Auto

It could have been really good, because the Resvent proved to be a stable and precise machine without leaks. Unfortunately, it lacked efficiency and with the leakage the limit turned out to be 11.4 cm H2O. Definitely not enough, but still a better result than in Prisma Smart, not to mention Apex.

Philips Dremstation GO

GO was bravely coping with it, but due to its compact size it doesn't stand a chance against bigger machines. The performance ends with a large leak at 10 cm H2O. For smaller leaks, the compensation works well. For a portable device it's really great!

ResMed AirMini

AirMini's not faking anything. This machine simply doesn't compensate for major leaks due to limited performance (was no problem with 20l/min.). These are the costs of compact dimensions. That's why ResMed gives you the option to use only selected masks. Without leaks, there are no problems with performance/ precision.


The first conclusion is that almost all devices do very well to maintain pressure precisely as long as there is no large leakage. In my opinion, the 3% limit in our test should be an acceptable level of accuracy. This means that the vast majority of CPAPs available on the market will provide an effective treatment while minimizing leakages. 

Only APEX iCH turned out to be a machine that stands out in this comparison. Bearing this in mind, I recommend caution also before buying unknown Chinese devices before checking the parameters of pressure in the service. The use of such a device may be not only inefficient, but also dangerous for health! 

We should positively evaluate Löwenstein, because it is very precise both in a cheap and more expensive device

Compensating for large leaks is a big challenge for CPAP devices. The problem is with both precision and performance. The absolute ideal in this respect is Prisma LAB. The agreed limit of "decency" set at 15% is maintained by Philips (the old Remstar is particularly positive), ResMed and DeVilbis. These are our finalists!

As you can see, I have removed the results of portable devices from the list of leakages, because I believe that it would not be fair to compare them in this way with fixed ones. Philips Dreamstation GO did better with two tested devices, but it can only compensate for pressure up to 10 cm/H20. The AirMini designers approached the issue in a different way and gave up compensation in general, but instead forced users to use masks with potentially low leakage. I can accept this approach, given that this is the smallest device on the market. 

Finally, a few more sentences of commentary on the test itself. Precision, pressure stability and performance are fundamental for any CPAP. Please note, however, that these are not the only criteria you should follow when buying a CPAP. Especially the algorithm and the comfort of the therapy are also important. If you want to know more about the devices presented here, we invite you to our purchase guide and tests of individual devices.

The undisputed winner of our test is Löwenstein Prisma LAB.

Brak komentarzy:

Obsługiwane przez usługę Blogger.