EERS Experiment Data (sherwoga)

[slowriter]

I can already report that the impact of changing the settings for EPR and EERS will be small and may not be statistically significant.

....

I am encouraged by what I see and believe I could improve my own sleep performance with my current equipment and the right size EERS ...

Aren't these two statements contradictory?

[sherwoga]

I can already report that the impact of changing the settings for EPR and EERS will be small and may not be statistically significant.

....

I am encouraged by what I see and believe I could improve my own sleep performance with my current equipment and the right size EERS ...

Aren't these two statements contradictory?

Thank you for the question and my answer would be, "Seems so".  But between the statements were thoughts about increasing the number of runs.  Doing so might improve the statistics enough to negate the first statement.  I hope so.  Regardless, I've learned a lot about my situation/treatment.  And that engenders the optimism of the 2nd statement.  

12/11/2019 EPR(Min Pressure),EERS pair = 3(8),0 Sequence #8 Design #2-1
The suffix (-1) on the design # is to indicate that this is a repeat to generate data to use to substitute for the outlier generated on 12/06/2019.  

I'll try to abstain from any further contradictory speculation and wait to see what the data says.  

           

[slowriter]

One issue not currently accounted for is the possibility of false centrals reported under the "clear airway" event categories. We really want to know what happens to the real central apneas with and without the EERS deadspace.

Won't impact data collection, but would on the reporting end.

[Gideon]

One issue not currently accounted for is the possibility of false centrals reported under the "clear airway" event categories. We really want to know what happens to the real central apneas with and without the EERS deadspace.

Won't impact data collection, but would on the reporting end.

To be clear, these are false Central SLEEP Apneas because of being in a waking or awake state.

[sherwoga]

12/12/2019 EPR(Min Pressure),EERS pair = 0(5),0

I have decided to collect four more nights of data to try to improve statistics.  This is the first of those four.

           

[JoeyWallaby]

Best of luck with your experiments sherwoga. I've given up on CPAP (changing to ASV) for now... I can get no/very minimal CAs if I don't treat FLs but once I use enough PS to treat FLs, I get CAs (the ouroboros of CPAP for me).

[sherwoga]

12/13/2019 EPR(Min Pressure),EERS pair = 0(5),18

           

[sherwoga]

12/14/2019 EPR(Min Pressure),EERS pair = 3(8),18 Sequence

           

[sherwoga]

12/15/2019 EPR(Min Pressure),EERS pair = 3(8),0

           

This run completes the set of 16.  As with other runs at this EPR and EERS setting combination, there is a significant cluster of CA events at the end of the night after I was awake.  I'm interested in hearing more discussion on the appropriateness of manually excluding these clusters on the basis that they are not truly SLEEP apneas, since I wasn't asleep.  It seems to me that there is something about this combination of settings causing the non-SLEEP apneas to occur since they don't occur at any of the other combinations.  Do I miss an opportunity to learn something about my body by simply ignoring them?

I will spend some time today beginning numerical analysis of the outputs.  Initially I will include all apneas, but I am fully aware that it might be better to exclude data per above.  

Tonight I plan on using a combination somewhere in the center of my design.  Find attached a graphical representation of the design.  I have four nights data at the four corners (marked with pluses).  I have no data for the ovals (blue or white).  The next step in the execution of a designed experiment would be to run the center point of the design.  As you can see in the figure, there is no true center point possible.  I would plan on testing some of the corners of the square in the middle of the figure instead.  After the numerical analysis, the best combination for testing may become obvious.  Absent those results, does anyone have any recommendation as to which of those four points I should test next?

   

[Gideon]

The SWJ events must be mentioned but I feel that without is valid. I would include them on your charts in red and footnoted

The interior points I would random select