Can a mask leak cause central AHI readings

Machine Resmed 10 Autoset

My mom has a weirdly high mask leak now — DME is sending out a new one. If someone has time to walk through the mask leak/AHI relationship or point me towards a resource that does, I’d be grateful.

Last night:
Leak is 23L/m

AHI is 15.3
Total AI 9.7
Central 8.8

I was surprised to see the central AI. How do mask leaks normally impact AHI? Do they just magnify what’s already there?

No SD Card in her machine but I’m have one today or Monday.

Centrals come from you holding your breath for more than 10 seconds with no airway blockage. It's common for them to occur is the person is partially awake while using the machine. I wouldn't be concerned with that amount of centrals. As far as the leak..The machine can compensate for leaks up to about 24L. If her leaks are consistently high they could come from mouth breathing with a nasal type mask. How is she sleeping?

kalatraza wrote: ↑

Sat Feb 25, 2023 11:10 am

No SD Card in her machine but I’m have one today or Monday.

After you get that, post one night's OSCAR chart in this thread. Without seeing the graphic data, you'll mostly get guesses about the issue at hand.

kalatraza wrote: ↑

Sat Feb 25, 2023 11:10 am

Total AI 9.7
Central 8.8

If that is the central index (8.8 centrals per hour), it's too high. The OSCAR graph should help us understand what is happening.

kalatraza wrote: ↑

Sat Feb 25, 2023 11:10 am

My mom has a weirdly high mask leak now

When having a discussion like this, please specify the mask model.

The CA index is higher than would be desirable. But it sounds as though this is an unusual reading for her; is that right? If so, just keep an eye on it.

The only connection between CAs and leaks that I can think of is this: CAs sometimes occur after a bit of arousal breathing -- the deeper breathing in the arousal washes out some CO2 from the bloodstream, which slightly delays the "breathe-now" signalling in the brain. If the pause between breaths is 10 seconds or more, the machine flags it as a CA. This could be connected with leaks if the leaks are causing arousals.

kalatraza wrote: ↑

Sat Feb 25, 2023 11:10 am

Can a mask leak cause central AHI readings

No.

kalatraza wrote: ↑

Sat Feb 25, 2023 11:10 am

Machine Resmed 10 Autoset

My mom has a weirdly high mask leak now — DME is sending out a new one.

Another thing to check is the hose itself. A few weeks back my husband had a pin prick sized hole in his hose that resulted in a near constant leak that was above his normal baseline leaks. And it was irritating to him; swapping the hose fixed the problem. Later on I did some fiddling with the hose to find that pin-hole puncture that created the leak.

If someone has time to walk through the mask leak/AHI relationship or point me towards a resource that does, I’d be grateful.

Last night:
Leak is 23L/m

This is a so-called 95% number. It means that your mom's leak rate was at or BELOW a level of 23 L/min for 95% of the time the machine was run during the reporting period. It also means that your mom's leak was ABOVE 23 L/min for no more than 5% of the run time during the reporting period.

As an example: If your mom used the machine for 8 hours, then her leak was at or BELOW 23 L/min for 7 hours and 36 minutes. If your mom had a leak that was 23L/min and the leak lasted 40 minutes and her leak for the rest of the night was 0 L/min, her 95% leak rate would be reported as 23 L/min.

It's worth pointing out that if you look on the machine's LCD or the MyAir app, and the overnight 95% leak rate is 23L/min, then Mr. Green Smiley Face most showed up in the morning: Mr. Red Frowny Face only shows up if the 70% leak rate is at least as 24 L/min.

To really understand what's going on with your mom's leaks you need to look at the data in Oscar or SleepHQ.

AHI is 15.3
Total AI 9.7
Central 8.8

I was surprised to see the central AI. How do mask leaks normally impact AHI? Do they just magnify what’s already there?

In general leaks don't affect the AHI unless they are so huge that the machine is having trouble maintaining pressure. For Resmed machines, the official information from the company indicates that unless leaks are above 24 L/min for at last at least 30% of the night, the leaks are not likely to have a significant affect on the efficacy of the therapy. Once the leaks get really large, machines can also have trouble measuring the airflow into and out of the lungs accurately, and when that happens, machines can fail to detect apneas or hypopneas. In other words, in the presence of a really horrendous leak, very real events might not be scored because the machine is having a tough time measuring the breathing in the first place.

Having said that, I want to also say this: A whole lot of people who use CPAPs will tell you that leaks that are less large and less long can still adversely affect sleep even if the machine has no trouble maintaining pressure. Leaks can wake you up and leaks can also cause arousals.

And the (normal) ragged breathing that can occur when you are drifting back and forth between brief wakes/arousals and very light sleep can, and often is, flagged as "events", even though these events would not be recorded on a full in-lab sleep study: The machine can't measure sleep stages and so it scores everything that meets its flow based definitions of OA, CA, and H, even if you are wide awake when using the machine. For some people, many or most of the fake events scored during very restless period or after a spontaneous arousal/wake are scored as CAs, but for others those fake events get scored as OAs or Hs. It just depends on how the breathing looks ragged.

So when leaks (large or small) are irritating enough to trigger restless sleeping, the overall AHI can go up. But the only way to really figure out if that's going on is to compare when the events are being flagged to when the leaks are occurring and to also look for evidence of less than sound sleep in the flow rate curve itself.

No SD Card in her machine but I’m have one today or Monday.

Once you have the SD card and some detailed data to share, it will be easier for us to help you help your mom.

Thanks everyone. I’ll check the hose and be back with moar data.