[Diagnosis] Small breaths?

[Michael2]

Hi:

I have been experimenting with the Lowenstein Prisma 30ST PAP machine for the last two months. I tried to upload my data to OSCAR and SleepHQ, but I encountered issues that suggest neither support the machine. Therefore, I attached a snapshot from last night generated by Lowenstein's own PrismaTS software.

In the snapshot from last night, you can see I take a small breath roughly 1 in every 10 breaths (marked by red arrows). Because the tops of the inhalations are ragged, my guess is that each small breath is caused by a severe obstruction that lasts for a single breath and so never registers as an apnea or a hypopnea. On a possibly related note, I sometimes wake up with a jolt as if my airway closed mid-inhalation. However, I do not recall this happening last night. 

I monitor my sleep by using the PrismaTS software to examine my breath-by-breath signal data, a camera to record my sleeping positions, and a Fitbit watch to track my sleep stages and SpO2 level. Sadly, I feel compelled to do all of this to resolve my many sleep apnea issues, and I long for the day when I can do none of it. I started seeing the aforementioned small breaths a few weeks ago. Last night, my EPAP was set to 16, and my PS was set to 4. However, I have seen the small breaths even after raising the EPAP to 20 and the PS to 5, suggesting they cannot be resolved by increasing the pressure. Moreover, I see the small breaths regardless of my sleep position, although they seem to be more frequent when I sleep on my back.

Do you guys have any thoughts on the small breaths? Do you think they are caused by obstructions? Could they be the result of epiglottis or palatal collapse? I will be seeing an ENT and undergoing a DISE in the near future, but I wanted to hear your thoughts on my PAP data.

Thanks,
Mike

[PeaceLoveAndPizza]

Looks like inspiration flow limitations. The graph shows the limitations as 20%, 40%, 10%, 20%, or thereabouts.

As you know, it is a fixed pressure machine, so it will not respond to the flow limitations.

[Michael2]

The Prisma 30ST PAP machine is not a fixed pressure machine. In fact, it has more functionality than virtually any other machine on market -- including timed-backup and ASV-like functions. I have attached a screenshot of the settings I used last night for which the screenshot of the respiratory flow was taken. It is a bit confusing because of Lowenstein's EPAP vs EEPAP. Effectively, the EPAP was set to 16, and the PS could vary between 4 and 10 to achieve a target volume of >=420 ml, but the PS remained at 4 throughout the night because the target volume did not go under 420 ml for a sufficiently long period of time. I have experimented with different EPAPs and PSs, and I can say that even an EPAP of 20 and a PS of 5 cannot eliminate the small breaths.

Anyways, for the purposes of this discussion, you can assume that the small breaths persist even with EPAPs up to 20 and PSs up to 5.

[PeaceLoveAndPizza]

Thanks for the correction on the machine, we do not see many of them. I was basing it on what you posted as that is the only information I had. The pressure was the same during the graph unaffected by the FL’s.

Regardless, the breaths are still flow limited. At those levels of flow limitations I would expect the machine to respond, but it did not. 

You can reduce flow limitations, but rarely completely do away with them. If it was supported by OSCAR it would be easier to see, but I suspect there are quite a few more. The goal is a 95% of <= 0.05. Not sure if your software shows that level of details.

[Macka]

Hi Micheal,
The machine should work with Oscar. Make sure you have v1.5.3

I am also using PrismaTS 5.13.0.13

Those breaths are smaller because the machine did not assist you that breath.
So we know the machine missed the trigger. Would have to check whether this was due to your Ti settings, or the little blips you have at zero giving false triggers etc.

[Michael2]

Thanks Macka. 

I will download the latest version of OSCAR. I will also experiment with Trigger In set to high sensitivity. In the past, when I had Trigger In set to high sensitivity, I would end up breathing so rapidly that it would be uncomfortable, but I will give it another try.

I am really impressed with Lowenstein. The 30ST is a beast with many treatment options, and the Prisma software is excellent.

[Macka]

Yes auto trigger may be getting messed up with your instability at zero. I take manual control of all those timings.

I have found adding more eepap can steady time at zero when the machine is watching for the trigger, so to speak. Have you tried bisoft 2 which gives 2cm eepap and can help stabilze that period?

[Michael2]

Yes auto trigger may be getting messed up with your instability at zero. I take manual control of all those timings.

I have found adding more eepap can steady time at zero when the machine is watching for the trigger, so to speak. Have you tried bisoft 2 which gives 2cm eepap and can help stabilze that period?

Thank you for your suggestion. Last night, I changed Trigger In to 2 (medium sensitivity) and Bi softPAP to Bi soft2 (EEPAP - EPAP = 2). The frequency of the small breaths dropped from roughly 1:10 for the previous night to 1:100 for last night. Moreover, my breathing stability index increased from 66 for the previous night to 96 for last night, indicating a dramatic improvement in the quality of my breathing. I will do more tests to see which of the two settings, Trigger In or Bi softPAP, has the greater impact on the frequency of the small breaths. When I set Trigger In to 1 (high sensitivity), I still found I breathed too rapidly, so I may need to keep it at 2 or Auto. 

In our previous discussions, you mentioned you slept better with the 25ST compared to the 30ST, and you speculated that could be due to the Ti/T control on the 25ST. Do you still feel you sleep better with the 25ST?

[Macka]

No, I am full-time on the 30ST now. IMO it can't be topped, with volume assurance on top of all bi-level modes. It just took a while to monitor what everything was affecting, and then learn to tune it.

I have a pretty good feel for setting these up now. I exclusively use bisoft2. The 2cm rise to EEPAP seems much more comfortable and workable for therapy, than off, 1 or 3 (tri-level) Off and 1 are too much like Resmed algorithm imo and tri-level 3cm rise to EEPAP is too much.

You must realise that all timing settings are first and foremost built around which of these bisoft levels you pick. Then external variables like vcom acting as a mechanical ramp to inhalation, hose and mask changing resistance and impacting inhalation as well.

Trigger IN 1 seems to work well on bisoft 2, if zero flow is stable enough.
Trigger EX seems to be best for most people on 2, if using bisoft 2. You have to check your flowsignal during the best periods, to see where the trigger is firing and that it is not causing any obvious problems.
Ramp IN I move to suit where peak pressure is arriving. This will be affected by all the variables previously mentioned, and so can only be set by looking at your inhalation signal once again. This is best studied in your best, flattest sections as well. Look for the little peak to be left, right or middle of the inhalation curve. 2 seems best for a lot of people.
Ramp EX seems best at 1 for everyone. For ex Resmed users, I think it may seem counterintuitive. I think it may be because speed of the exramp causes a lot of complaints on Resmeds. The very late arrival of peak inhalation pressure means they have to steepen the exramp to get down to epap intime. Despite this, they still don't get there till after exhalation peak. I find this pretty unsettling and the reason for a lot of unstable zero time in some Resmed users.

The Loewenstein starts earlier and arrives at epap well before peak exhalation. It is very steadying if pressure stabilises during peak exhalation imo.

The volume assurance set 80ml below your ideal also seems to be pretty good for most people. I set the response to slow. This means it will stay very stable, but start to respond gently when flow limitations start to persist and also leave the extra pressure support in place for a while. I found medium and fast would respond, add the extra PS then drop again too soon, after breathing shapes were better, and have to just do it all again. I find slow will keep the extra PS in place for larger durations of flow limitation like during REM when your breathing is waxing and waning the whole time. I would rather pressure support go from 5 to 6 and stay there for REM dropping back to 5 in the next deep sleep period and being stable there. This is what I aim for it to look like below.

   

Remember also that your pressure support or PDIFF, is made up of the rise to EEPAP, then the rise to IPAP. So PDIFF 5 uses 2cm in the EEPAP rise, and only has 3cm left for IPAP. You will end up using a seemingly higher PS then you would on a Resmed, or other device. I know you probably get how this works already, but thought I would just mention it anyway. You need to get the Resmed standards for PS out of your head when EEPAP is chewing up some of your pressure support. 5, 6 & 7cm will be much more common with bi-soft 2. That effectively will give you true pressure support of 3, 4 & 5cm. I use PDIFF 5-6.5, which is equal to 4-5 on a Resmed.

The deepsleep indicator and sleep stability score, are two of your best assets for assessing the real effect of changes. You start to get good therapy at 85 stability, & 100min deep sleep. Great therapy starts at 90+ stability and 150+min deep sleep. Around 95-96 is really good.

You do need a combo of Oscar and PrismaTS data to properly track your progress.

[Michael2]

I have yet to experiment with vcom with the 30ST. I will give it a try soon.

I am also using prismaTS [Ver: 5.13.0.13], but I do not see any deep sleep indicator. How do I view it?