ResScan UnderReporting or SmartLink OverReporting or ?

[DHC]

Using both the IntelliPAP AutoAdjust and the S9 AutoSet for a while now, and alternating usage to get an idea of the differences between machines and reports. I am keeping the settings identical between the two as much as possible - so the min/max pressures, ramp time and start are the same. I *did* make an upward adjustment in the min/max settings to both machines - raising the min to 6.0 and the max to 12.0. As best I can tell both devices and software are using the same parameters for event detection, hence, that should not be causing any differences.

Anyway - here is a table of results - just showing AHI for now. ResScan consistently reports MUCH lower AHI. That would be a good thing *IF* it can be believed. Since SmartLink reports considerably higher AHI, it calls BOTH into question. My numbers even on the high side do not seem to be a serious concern, but the large differences in reported AHI between the two devices is at least curious.

Sep 27/28 - AHI 1.33 (IP)
Sep 28/29 - AHI 2.86 (IP)
Sep 29/30 - AHI 0.6 (S9)
Sep 30/Oct 01 - AHI 0.3 (S9)
Oct 02/03 - AHI 0.2 (S9)
Oct 03/04 - AHI 0.7 (S9)
Oct 04/05 - AHI 5.19 (IP)
Oct 05/06 - AHI 3.43 (IP)
Oct 06/07 - AHI 4.03 (IP)
Oct 07/08 - AHI 3.25 (IP)
Oct 08/09 - AHI 3.55 (IP)
Oct 09/10 - AHI 0.2 (S9)
Oct 10/11 - AHI 0.2 (S9)
Oct 11/12 - AHI 0.3 (S9)

A while back I watched the excellent DeVilbiss tutorials on their process for identifying and responding to events. It inspired confidence that DeVilbiss has it 'right' - which makes me concerned that ResScan is UNDER-reporting events.

Any thoughts on how to determine which of the two (or neither) are accurate?

[jnk]

The clinicical guide for my machine says the following:

"The AHI values reported by the VPAP Auto should be viewed as trending information only, that may suggest the need to further investigate the patient's condition with a polygraphy study. . . . The AHI reported by the VPAP Auto may be higher than the AHI determined by polygraphy scoring since the VPAP Auto cannot detect the sleep state or the presence of arousals, nor does it incorporate SpO2 measurements into the AHI calculation."

My conclusion from the above is that one cannot successfully compare the self-reported estimates of events across different brands of machine, or, for that matter, even across different series of the same brand, in the case of S8s (like the VPAP Auto) and the S9 series. Those estimated "AHI" numbers are given only because they are useful for assessing whether therapy may be improving or deteriorating over time for the one patient using the same machine every night, but making comparisons from machine to machine in order to compare efficacy is NOT the purpose of such data, as I understand it. And no machine manufacturer would be foolish enough to claim that their estimates are equivalent to the AHI found in a PSG, as ResMed made so clear in the statement quoted above.

[carbonman]

Sep 27/28 - AHI 1.33 (IP)
Sep 28/29 - AHI 2.86 (IP)
Sep 29/30 - AHI 0.6 (S9)
Sep 30/Oct 01 - AHI 0.3 (S9)
Oct 02/03 - AHI 0.2 (S9)
Oct 03/04 - AHI 0.7 (S9)
Oct 04/05 - AHI 5.19 (IP)
Oct 05/06 - AHI 3.43 (IP)
Oct 06/07 - AHI 4.03 (IP)
Oct 07/08 - AHI 3.25 (IP)
Oct 08/09 - AHI 3.55 (IP)
Oct 09/10 - AHI 0.2 (S9)
Oct 10/11 - AHI 0.2 (S9)
Oct 11/12 - AHI 0.3 (S9)

Veeerrry intaaresting!

[jonquiljo]

"The AHI values reported by the VPAP Auto should be viewed as trending information only, that may suggest the need to further investigate the patient's condition with a polygraphy study. . . . The AHI reported by the VPAP Auto may be higher than the AHI determined by polygraphy scoring since the VPAP Auto cannot detect the sleep state or the presence of arousals, nor does it incorporate SpO2 measurements into the AHI calculation."

Even without knowing sleep state or the presence of arousals these machines should (if their reporting capabilities are worth anything) get the number of sleep related breathing events and calculate an AHI within reasonable proximity to each other. This wording sounds like legal CYA to justify bad reporting.

It's pretty damn scary that the numbers are that far off! You're showing numbers that indicate as much as a 10-fold difference in sensitivity! These machines should either measure events or not pretend to tell us the truth. Isn't the "automatic" function on APAPs all about detecting events and trying to compensate? If it cannot reliably detect events then it cannot reliably compensate either. I guess this is all the more reason to incorporate O2 monitoring at home (on occasion) to verify your therapy.

The bottom line is that different brands of machines should be closer in reporting that that if they are to be taken seriously at all. Of course it's not a sleep study. But these machines can either detect events or not. If they really can't - they shouldn't pretend to.

[ozij]

The machine's don't give you the same therapy, therefore you dont' get same results.

The DeVilbiss Intellipap only respond to snores, apneas and hypopneas. It does not use any other aspect of your breathing as a reason to chang pressure, and it responds to apneas and hypopneas identically.

The ResMes is far more sophisticated in its analysis, and uses in addition, flow limitations in your breathing to change its response. It has a differential response to diffent breathing parameters, and, surprising as that may seem, it does not change pressure when it encouters what it deems a hypopnea.

I would say your result are proof positive that the ResMed is giving you better therapy.

O,

[jnk]

. . . different brands of machines should be closer in reporting than that if they are to be taken seriously at all. . . .

I agree 100% that it would sure be nice if they all used similar definitions. That would be better for patients. But maybe not so good for manufacturers (?) .

In the meantime, all we can do is to keep in mind that there are differences in how the events are calculated and scored, which is, I assume, because of proprietary scoring algorithms based on partial hypopnea-defining data, as illuminated by statements from companies about some of their machines, as documented, for example, in the famed Velbor Chart of Differences:



Is the info out on how the S9s calculate and score, I wonder?

[carbonman]

Any thoughts on how to determine which of the two (or neither) are accurate?

So....do you feel any difference from
using one or the other machine??

I would love to get my grimy mitts on an S9 for a month.

[Emilia]

What were the 90th %tile pressures for those readings? That may shed some light on how the machines were responding and providing the appropriate pressure within that range. It is very interesting data. I was titrated at 8 but using 7-10 as a range in my IP Auto, my 90th %ytile rarely goes above 7 with a max, on rare occasions, of 8. My AHI is consistently below 1.5 and many nights below 1. I found when I started on the IP, with a range of 6-10, my AHI was up around 3-5 so bringing up the lower pressure closer to my titrated pressure seemed to help immensely.

Perhaps, DeVilbiss Marketing can shed some light on your findings...... I suspect she'll be along soon to comment!

[jnk]

I would love to get my grimy mitts on an S9 for a month.

How much do ya want for one of those AutoSets sitting in your closet?

[DHC]

The machine's don't give you the same therapy, therefore you dont' get same results.

The DeVilbiss Intellipap only respond to snores, apneas and hypopneas. It does not use any other aspect of your breathing as a reason to chang pressure, and it responds to apneas and hypopneas identically.

The ResMes is far more sophisticated in its analysis, and uses in addition, flow limitations in your breathing to change its response. It has a differential response to diffent breathing parameters, and, surprising as that may seem, it does not change pressure when it encouters what it deems a hypopnea.

I would say your result are proof positive that the ResMed is giving you better therapy.

O,

I certainly considered the QUESTION of whether the S9 is providing improved therapy. When I look at the charts for Pressure and Events, both devices show a VERY similar profile in terms of responding to events.

Your comment about the IntelliPAP only responding to snores, apneas and hypopneas is interesting. The DeVilbiss webinars described their product as "predicting" and responding to respiratory events by sensing apnea and hypopnea events (defined with much different parameters), detecting leaks, detecting expiratory puffing, and detecting central apneas (they use the term "non-responding event") - with the goal being to reduce the AHI.

I have not been able to find very much about the ResMed algorithm, though I do understand they rely on flow limitation in addition to snoring in their predictive algorithm. I also noted that ResMed says they respond to events above 10cmH2O - which would seem to be pretty seldom in my case - per the following table for 95th Percentile Pressure:

Sep 27/28 - 4.0 (IP)
Sep 28/29 - 8.0 (IP)
Sep 29/30 - 9.1 (S9)
Sep 30/Oct 01 - 8.9 (S9)
Oct 02/03 - 9.9 (S9)
Oct 03/04 - 9.6 (S9)
Oct 04/05 - 7.5 (IP)
Oct 05/06 - 7.5 (IP)
Oct 06/07 - 7.0 (IP)
Oct 07/08 - 6.5 (IP)
Oct 08/09 - 7.0 (IP)
Oct 09/10 - 9.9 (S9)
Oct 10/11 - 10.2 (S9)
Oct 11/12 - 9.6 (S9)

I am just not too sure that the answer for the differences is that the S9 is providing superior therapy. It *may* be the case - and if so, would DEFINITELY be worth knowing.

Is there any other data from ResScan and/or SmartLink that might be helpful? I can post some graphs if you can point me to the ones that should be posted.

[jdm2857]

The machine's don't give you the same therapy, therefore you dont' get same results.

The DeVilbiss Intellipap only respond to snores, apneas and hypopneas. It does not use any other aspect of your breathing as a reason to chang pressure, and it responds to apneas and hypopneas identically.

The ResMes is far more sophisticated in its analysis, and uses in addition, flow limitations in your breathing to change its response. It has a differential response to diffent breathing parameters, and, surprising as that may seem, it does not change pressure when it encouters what it deems a hypopnea.

I would say your result are proof positive that the ResMed is giving you better therapy.

O,

Maybe you should try straight CPAP (at your typical 90- or 95-centile pressure) on both machines to factor out the effects of the auto algorithm. You'll get a much better picture of the machines' event detection abilities.

[DHC]

Maybe you should try straight CPAP (at your typical 90- or 95-centile pressure) on both machines to factor out the effects of the auto algorithm. You'll get a much better picture of the machines' event detection abilities.

Yes, that should remove the response algorithm from the equation - provided the parameters for definition of the events is the same.

It looks like the S8 AutoSet defines them as:

Apnea - 75 % decrease in flow for a period of 10 seconds (simplifying the chart upthread)
Hypopnea - 50 % decrease in flow for a period of 10 seconds

DeVilbiss defines them thusly (again simplifying):

Apnea - 90 % decrease in flow for a period of 10 seconds
Hypopnea - 50 % decrease in flow for a period of 10 seconds

Assuming the S9 uses the same definition as the S8, it would seem the ResScan results would show MORE events - apneas, in particular, than SmartLink.

Alternatively, the IP allows for changes in the event detection parameters, so I suppose I could reset them based on the ResMed criteria and see what results.

As a point of reference, the CPAP Wiki offers these definitions:

Apnea is the cessation of airflow for 10 seconds or greater.

Hypopnea: A respiratory episode where there is partial obstruction of the airway lasting greater than 10 seconds. Also called partial apnea or hypo-apnea.

CPAP Wiki is notably silent on the specific parameters to be used for the definition.

[carbonman]

Any thoughts on how to determine which of the two (or neither) are accurate?

So....do you feel any difference from
using one or the other machine??

???????????????????????

I would love to get my grimy mitts on an S9 for a month.

How much do ya want for one of those AutoSets sitting in your closet?

That would be a tough one.
A proven S8 for an S9.....hummmmm.



So, with these S8 numbers for a year,
I should be able to pull 0.0 AHI on a nightly basis w/an S9???

At this point in time, I'll keep what is tried and true.

[DHC]

Any thoughts on how to determine which of the two (or neither) are accurate?

So....do you feel any difference from
using one or the other machine??

???????????????????????

I did not answer it earlier because, candidly, I am unsure. You see, my need for CPAP came about as the result of a cervical spine fusion surgery. The surgery has led to any number of issues and how I "feel" is, perhaps, the LEAST reliable indicator at the moment. The surgeon recently referred me to a "high-end ENT" whom I visited today and learned that I will need yet another surgery. This ENT also told me that "right now" the use of CPAP is "critical" based on his examination and findings. Based on recommendations, past experience with ENT docs who were FAR more about cosmetic stuff - AND the nature and findings of the exam today, I think this guy is 'bang on' - though I *HATE* the idea of another surgery so soon.

Anyway - the short answer to your question is that I feel generally the same with both machines. If anything, the increased pressure of the S9 causes me a few more mask leak problems leading to waking during the night.

[carbonman]

I did not answer it earlier because, candidly, I am unsure. You see, my need for CPAP came about as the result of a cervical spine fusion surgery. The surgery has led to any number of issues and how I "feel" is, perhaps, the LEAST reliable indicator at the moment.
Anyway - the short answer to your question is that I feel generally the same with both machines. If anything, the increased pressure of the S9 causes me a few more mask leak problems leading to waking during the night.

Good point. Thanks for the clarification.