Recent switch to BiPAP -- need help w/ flow rate abnormalities and settings

[mcrick]

Hi all,

My overall goal is to find the lowest-stress pressure set-up that flattens those UARS-style flow limits so I finally wake up feeling rested.

I ran my AirCurve 11 VAuto last night (only my 4th night on the machine since swapping from CPAP) purely so the machine would log flow-limitation data, but I fixed pressures so it behaved exactly like S-mode (IPAP 13 / EPAP 7, PS 6).  Aerophagia was mild (tolerable) at these pressures, and AHI came in at 1.48 (mostly flagged CAs with scattered OAs).  What puzzled me were a series of “flat” or ragged inhalations.  Glasgow Index overall = 1.12 with Variable Amplitude doing damage (0.38) plus Skew (0.40) and Top-Heavy (0.17).

Links below:
 • All nine cropped OSCAR zoom-ins covering each flagged event (and a couple of unflagged disturbances) and Glasgow Index 
 • SleepHQ dashboard here

My main ask: How would you interpret the disordered flow-rate shapes?  Are they classic UARS flow-limits, “high loop-gain” wiggles, or just normal recovery breaths?  And are Variable Amplitude or Skew on the Glasgow Index worth chasing with settings changes?

Background & context:

• Goals: Minimize flow limitations / UARS arousals first, AHI second.  AHI < 2 is typical; I still wake groggy unless the FL channel is nearly flat.
  
• Body position: Mostly side sleeper but roll supine; side-sleep “trainer” sometimes used but fragmenting sleep when I change positions / roll over.
  
• Centrals: Not historically an issue; last night’s CAs might be scoring artefacts or sigh-pause rebounds.
  
• Recent settings experiments (3-night plan):  
      • Night 1 – 12/7 S-mode, TiMax 3.4 s, TiMin 0.30 s, Trigger High, Cycle Low → Glasgow 1.24, AHI 0.85, zero aerophagia  
      • Night 2 – 13/7 (same timing & trigger/cycle) → mild aerophagia, Glasgow 1.12, AHI 1.48 (data below)  
      • Night 3 (tonight) might be 12/6 with same timings to see if Variable-Amp settles.
  
• Why Glasgow Index? I coded a script to mass-correlate Glasgow, AHI, FL, EPAP, IPAP & PS across ~250 CPAP nights; PS/EPR shows the strongest negative correlation with Glasgow (-0.66).  EPAP correlates positively (+0.49), so I’m walking a fine line between splinting the airway and avoiding aerophagia.
  

Device settings used last night:

Code:

Mode          : VAuto (Min EPAP = Max EPAP = 7)
PS            : 6  (fixed)
Max IPAP      : 13
Trigger       : High
Cycle         : Low
TiMin / TiMax : 0.30 s / 3.40 s
Mask          : FFM
Ramp / SmartStart : Off
Humidifier    : Manual, level 3

What I’m looking for:

• Are the “flat” stretches true flow-limits that my machine isn’t catching?  
  
• Does Variable Amplitude (Glasgow 0.38) signify unstable drive or something else?  
  
• Would dropping TiMax further (e.g., 3.0 s) or raising Trigger to Very High smooth these? I find that Ti Max under 3.4s cuts off my breaths when trying to fall asleep.
  
• Any other tweaks to test while staying ≤ 7-8 EPAP so aerophagia doesn’t worsen?
  

About me (medical & PAP background):

• Diagnosed two years ago with mild-to-moderate obstructive sleep-apnea (sleep-lab AHI ≈ 10) but symptoms point strongly to UARS / flow-limitation–driven arousals: unrefreshing sleep, daytime fatigue, and “choking/sigh” awakenings even when event index is low.
  
• No significant cardiac, pulmonary or neurologic disease; BMI normal, active mid-20s male, 6'3". Narrow palate and large tongue (malampati class 3 or 4) but no major nasal obstruction. Previously had successful functional septorhinoplasty with septum correction and turbinate reduction.
  
• P30i large mask + mouth-taping + mandibular advancement device (MAD) is my usual nightly toolkit.
  
• Machine: ResMed AirCurve 11 VAuto since 4 days ago (previously AirSense 11). Use OSCAR and SleepHQ for data, plus a self-written script that calculates and graphs the Glasgow Index from EDF files so I can quantify flow-shape abnormalities.
  
• Typical therapy metrics on fixed-pressure bilevel:
  
• AHI 0.5 – 2.0 (usually mostly clear-airway flags)
  
• Flow-Limitation (ResMed) ≈ 0 but Glasgow Index hovers 1.2 – 1.8.
  
• Aerophagia: happens reliably when EPAP ≥ 9 cmH₂O, seemingly regardless of IPAP; milder or absent at 7 cmH₂O. Wedge pillow, side-sleeping bumper belt, and MAD all help but don’t eliminate it.
  
• Centrals haven’t been a real issue historically; scattered CA flags look more like post-sigh pauses.
  

Really appreciate any insights, especially from folks who’ve chased UARS-style flow-limits on bilevel!
I've been exhausted every morning when I wake up with CPAP and bilevel so I am just looking for any ideas.

Thanks!

[mcrick]

https://imgur.com/a/MdmjvDV

Above are last night's results with the settings below:

• Mode VPAPauto
  
• Min EPAP 5.20 cmH2O
  
• Max IPAP 11.60 cmH2O
  
• PS 6.40 cmH2O
  
• Climate Control Manual
  
• Cycle Low
  
• Humidifier Status On
  
• Humidity Level 4
  
• Mask Full Face
  
• Ramp Off
  
• Smart Start Off
  
• Temperature 29 ºC
  
• TiMax 3.40 Seconds
  
• TiMin 0.30 Seconds
  
• Trigger Very High
  

Glasgow index improved but aerophagia worsened (maybe a 5 or 6 out of 10) -- I woke up several times and burped a lot in the morning. I think the aerophagia offset the improvements in glasgow because I feel tired again today. I don't know if it was the EPR, IPR, or PS, or other (maybe trigger very high or Ti max is too much). It seems that glasgow index keeps improving the more I increase PS, but I don't know if it was the PS or the IPR or something else that caused aerophagia. I have historical evidence to suggest that raising IPR worsens aerophagia, but I don't have enough data to know how PS factors in, other than knowing that PS of 0 or 1 makes it much worse.

I'd greatly appreciate any help with making sense of all this data.

[ahuman]

Perhaps @DaveSkvn can chime in with more details about the Glasgow index. Possibly one aspect of ASV that might work for some is PS can be higher when needed rather than all night which can cause aerophagia.

[mcrick]

   
https://imgur.com/a/OkxWMFm

I re-ran my custom analysis software which compares CPAP data to glasgow index and had it take in all metrics that are computed in Dave's Glasgow Index website.
The results are interesting but I can't quite make sense of them.

Raising PS/EPR helps almost every glasgow index metric considerably, except it worsens flow rates with flat, heavy tops.
The only thing that seems to somewhat help those two metrics are raising EPAP and IPAP, but doing so worsens my aerophagia.
Raising PS/EPR above 5ish also seems to worsen aerophagia.

So now I am stuck trying to figure out how to minimize all of these metrics when they are improved by opposing changes in settings. 

[mcrick]

Bump.

I’ve uploaded my most recent AirCurve 11 + O2 ring data to my SleepHQ dashboard and would really value this forum's take on it; I am really feeling the effects of the subpar sleep quality.

What I’m seeing so far:

• Combining the BiPAP with my mandibular device has given me a few OK nights (inconsistently), but whenever I push pressure-support above ~6 cm the AHI usually climbs—mostly from clear-airway flags. I’m not sure how clinically meaningful these CA events are.
  
• EPAP above ~6 cm (especially with PS > 6 cm and a very-high trigger setting) seems to bring on aerophagia.
  
• I’m still waking up groggy, with puffy nasal tissues and mild congestion.
  

What I’d love your help with:

• Interpreting the flow-rate shape and the CA clusters—do they point to CO₂ wash-out or something else?
  
• Pinpointing settings that could smooth the flow curve without driving aerophagia or CAI.
  
• Whether a step down in humidity/tube temp might relieve the congestion without hurting comfort.
  
• Reducing sleep fragmentation in general. My Apple watch data (as seen on the SleepHQ dashboard) seems to hint at lack of deep sleep with a lot of fragmented stages and bouncing between stages sporadically.
  

I realize my frequent setting changes make the picture messier, but any help in reading these traces would be much appreciated. Let me know if you need anything else from me.
Thanks a lot for taking a look.

P.S. I nerded out and wrote code to analyze my PAP data, run it through the Glasgow Index website, and analyze correlations. It might mean nothing, and it's only trained on my 14 nights of bipap data so far, but I attached the correlation matrix if you find it interesting or helpful.