large leak S9 VPAP Auto

I just changed over from a S9 AutoSet to this machine and have been on it for 3 nights. Since the first night I've had large leaks in my sleep report reading on ResScan software. This morning my AHI was 4 and the leak rate read from the sleep report on the machine was 31 L/M. When I pressed the i bar this morning my usage was 6.9, Mask had a yellow smiley and my AHI was 4. How can you have an AHI of 4 with that large of leak? When using the AutoSet my leaks were usually under 7 mostly zero. I'm using all the same equipment and the only thing that's different is the blower part. I'm starting to think there might be a leak in it. Any ideas. I've put another climate line on for tonight.

squid13 wrote:I just changed over from a S9 AutoSet to this machine and have been on it for 3 nights. Since the first night I've had large leaks in my sleep report reading on ResScan software.

Can you tell us exactly what the settings on both the old S9 AutoSet were and what the settings on the new S9 VPAP Auto are? That's important info for trying to understand what might be going on.

This morning my AHI was 4 and the leak rate read from the sleep report on the machine was 31 L/M. When I pressed the i bar this morning my usage was 6.9, Mask had a yellow smiley and my AHI was 4. How can you have an AHI of 4 with that large of leak?

The short answer to your question is that your machine detected something like 27 or 28 events during the night that met its algorithm's criteria for being scored as an apnea or a hypopnea. Whether all of these events are "real" or not may be harder to determine. But over the course of the 6.9 hour night, there was plenty of time where the leak rate was likely to be low enough to record reasonably accurate data.

How might that happen?

First, that 31 L/min leak figure is the 95% unintentional leak figure. Since you report usage of 6.9 hours, that means that you were leaking ABOVE 31 L/min for no more than 5% of the night or:

(3 minutes per hour)*(6.9 hours) = 20.7 minutes during the night

During the remaining 6 hours and 33 minutes of time you had the machine running your leak rate was AT or BELOW 31 L/min. And it's possible that during most of that 6:33, your leak rate was substantially below 31 L/min. Without looking at the data on the computer, there's no real way to determine if you simply had one bad 20 minute leak in an otherwise "normal" night as far as leaks are concerned or if you had substantial leaking for most of the night. You say a yellow Smiley face showed up on the on-board sleep report. I haven't hear of yellow Smiley faces. I've heard only green Smileys and red Frownies. Mr. Red Frowny Face shows up when the 70% unintentional leak rate is AT or ABOVE 24 L/min. (See Mr Red Frowny Face vs Mr Green Smiley Face on the S9.) Since you didn't see Mr. Red Frowny Face, that means that your 70% leak rate was below 24 L/min. Hence an upper bound on leaks large enough to affect therapy in the opinion of ResMed engineers is:

.3*6.9 hours = 2.07 hours = 2 hours and 4.2 minutes (i.e. 2:04)

And this 2 hour maximum for really bad leaks includes the 21 minutes where your leak rate was AT or ABOVE 31 L/min.

Now to further address your real question: "How can you have an AHI of 4 with that large of leak?"

Well, for at least 4 hours during the night, your leak rate was sufficiently low for the Resmed engineers to believe the machine was recording reliable data. All it would take is for you to have somewhere around 27--30 "reliable" events during those 4 hours with reliable data to give you an overall AHI = 4 for the whole night even if the machine was not able to detect a single event during the two hour period where you might have been leaking at a rate large enough to cause problems with data reliability in the eyes of the Resmed engineers.

Also keep in mind that once the leaks get very large and long, the data is unreliable, but does not mean the data is nonexistent. In other words, the machine continues to try to detect apneas and hypopneas even in the presence of large leaks. But because of the large leaks, the machine might detect more events that are actually there OR it could fail to detect real events OR both. We just don't know which way the errors in the data might crop up in the presence of large leaks.

When using the AutoSet my leaks were usually under 7 mostly zero. I'm using all the same equipment and the only thing that's different is the blower part. I'm starting to think there might be a leak in it. Any ideas. I've put another climate line on for tonight.

Where might the "new" leak be?

• Mask and hose leaks? How old is the mask and the mask cushion/nasal pillows? Higher than normal leaks can indicate that the mask and/or its cushion need to be replaced. Did you get a fresh hose with the new machine? If not, how old is the hose you are using? The climateline hoses do seem to be fragile for some people and they seem prone to developing leaks right at the coupling.
• and potentially higher IPAP pressure setting. Because the IPAP-EPAP difference is often much greater than the maximum 3cm drop on the S9 Autoset with EPR = 3, it could be that you need to carefully refit the mask and the headgear at your new maximum IPAP pressure. If IPAP - EPAP is substantially larger than 3, it could simply be that you have to learn to adjust the mask for the more constantly changing pressure between IPAP and EPAP.
• Mouth leaks? It could be that you are simply having some additional difficulty in keeping your mouth shut if your IPAP is higher than what your max pressure setting was on the Autoset. In other words, it could be that the switch from EPAP to IPAP is causing your mouth to open up on many inhalations

The pressure when using the AutoSet was 7 and 10, the pressure on the VPAP is 14 and 10. The mask I was using was a Swift Lt which checked OK on Mask check when going to bed and when getting up this morning. It's new, been worn twice before with no leak problems.The climate line was new but couldn't find any leaks in it so I put my old one on for tonight. My mouth was taped shut so it had no leaks. The smiley is green on mask check but when checking with the i bar he looks more yellow than green. If you look real close you can see a green tint above his mouth.So he's sort of a yellowish green. I also took the units apart and reseated them back together again. here is a graph of last night. I'm using straight bi level.

Just make sure the Mask cushion is properly clipped in. I had this happen to me once and I thought the machine was busted and it turned out to be the simplest thing.

The mask checks OK for all parts being installed right.

Those leaks are definitely intermittent in the sense that you are not leaking at or above 24 L/min all night long. Indeed, there's couple of nice hour long or more stretches with very low (or even 0) unintentional leak rates, which is not what I'd expect if there was a problem with the machine or the hose.

You said you taped your mouth, so you don't think that's where the leak is. Any chance your lips opened up even just a smidge with your tongue NOT on the roof of your mouth? For some folks, even the smallest lip opening combined with a bad tongue position can lead to leaks. Given that you are using quite a bit more pressure with the BiPAP than you did with the Auoset, it's possible that the additional pressure is much more capable of forcing a small gap in your lips to open when the tongue is not on the roof of your mouth. When the tongue is resting on the bottom of your mouth instead of resting pushed up against the roof of your mouth, that allows air to enter the mouth from the back of the throat. Going from an Auto range of 7--10cm to a fixed BiPAP of 14/10 means that there's a whole lot more air that can get into the mouth that way. And more air in the mouth means a higher chance that the air will be able to force open a small gap between your lips even if they are taped.

Or, since the Swift FX mask is a nasal pillows mask, any chance one of the inner cones collapsed for a while? That can lead to leaks when it happens to me. Any chance that you simply turned over into a not so great position as far as the mask was concerned? What I mean are these two things:

• Some folks find the small, heavy hose is prone to pulling the FX's pillows away from the nostril when they are sleeping in certain positions. Do you hang the hose? Do you "cuddle" the hose? Or is the hose just lying on top of the bed covers?
• Some people find that when they are sleeping on their side or stomach, their bed pillow can dislodge one (or both) of the nasal pillows enough to create a leak. And then it's pretty easy to reseal the mask in your sleep or when you're only semi-awake when you turn back into a sleep position that is less likely to smoosh the pillows out of place.

I use a hose buddy for my hose. I'm a left side sleeper and I don't toss a turn during my sleep. I pretty well stay put all night in one spot. I taped my mouth to see if that was it for the higher pressure. I used 2" medical tape and it was sealed tight cause when I take it off in the morning it feels like it's peeling my face off. I don't believe the air passes through the tape. When I had my S9 AutoSet I didn't tape cause my mouth stayed closed and tongue was in the proper place, my leaks were 7 or under usually zero. I used my Swift FX last (Mask most used) with new pillows and they stayed seated all night. Here is last nights graph.