[Treatment] Is AHI misleading?

[zimlich]

I believe AHI is not the best indicator of successful therapy, but it's the best we have now. I have a high AHI- average of 11 on ASV, but do not desaturate, most of my events are hypopneas and I basically feel ok. I used to take Provigil, but for some reason at the beginning of the year I stopped needing medicine to stay awake. If you're not consistently up to par you can see your sleep doc for a review of your reports. You can also get the pulse oximeter- I understand that if you're having events and not desaturating they are not damaging to the heart( per my sleep doc). Your events seemed to be in two main clusters during the night. I am wondering about a Zeo...it should give more information about your sleep stages during the night. Or there are iPhone aps. I am not suggesting you buy anything and have no vested interest in either company. We are all in different stages of health, but if you are of an age where you would normally feel pretty good and have plenty of endergy and this is not the case I would do more problem solving. Good luck, keep us posted.

[PaulaO2]

.... Several of us have the CMS-50D which is available from Supplier 19.

I believe Paula means the CMS-50D Plus, which is the non-medical device. The CMS-50D is the FDA approved item, which is more expensive. I think.

Duh. Yeah, that's what I meant. I was testing your reflexes. Yeah, that's what I was doing.

[SuperSleeper]

I do understand that reliance solely upon AHI to guage therapy effectiveness may not be the absolute best course of action.

HOWEVER, I still believe that AHI (for trend analysis) is the best single point of reference for determining whether or not our CPAP therapy is helping us, and the best way to determine whether or not setting and pressure changes help or hinder our therapy.

I'm not saying we need to discount the "how do I feel" method. Not at all. But we also have to keep in perspective that the "how do I feel" method is VERY subjective in nature. YES, You may "feel better" now that you've increased your pressure or other settings on your CPAP machine, but how do you know for certain that you're not feeling better because of a completely unrelated factor that you did not consider? You may be feeling better because you've gotten more exercise lately, or your diet has improved, you're taking a different vitamin or pill now, or you're drinking more liquids lately, or you've gotten a new more comfortable pillow or bed, you've got less stress this week, or it's not as hot this week as it was last week, etc., etc.

The AHI is not subjective at all. It is a hard data-point that simply reports your events. We know it's not a perfect method, but when your AHI goes up or down, that's a cold hard fact that you can use to scientifically analyze your overall trending data. Again, it's supposed to be used for trend analysis only. Different machines record AHI differently.... so when using AHI, you need to stick to one machine only when using your AHI data to determine therapy effectiveness over a long period of time.

When I judge my own CPAP's effectiveness, I think I generally weight my AHI as representing 80% of what's actually happening and the "how do I feel" method somewhere around 20% (because of it's very subjective nature). If my AHI is very good, but I feel lousy, I look at other things in my life that may be negatively affecting my health, since the AHI numbers look good, most likely my feeling lousy isn't due to an improper CPAP setting.

Also, for those who regard the "how do I feel" method as the best primary method for determining CPAP effectiveness, they must also realize that if the person is a relatively new CPAP user, they aren't necessarily going to "feel great" immediately. It can take weeks and weeks to acclimate to CPAP to where you start feeling better. In those cases, "how do I feel" will not give you a good indication of therapy effectiveness AT ALL.

Still think AHI is the single best method available for us to use, assuming we have a data-capable CPAP machine.

Just some thoughts. Your mileage may very.

[PaulaO2]

The AHI is not subjective at all. It is a hard data-point that simply reports your events. We know it's not a perfect method, but when your AHI goes up or down, that's a cold hard fact that you can use to scientifically analyze your overall trending data. Again, it's supposed to be used for trend analysis only. Different machines record AHI differently.... so when using AHI, you need to stick to one machine only when using your AHI data to determine therapy effectiveness over a long period of time.

And see, if this fact is to be so important--and I agree it is--why should/would the machines record it differently? Users would not tolerate a glucose meter that is off. That company would be out of business. Why do we tolerate machines that record such an allegedly important number as the AHI differently? If one machine says my AHI is below 5 and another says it is above 5, which is right? Both? Neither?

One of these days, xPAP users are going to get as organized as those with diabetes. I think that idea scares xPAP manufacturers.

[b360155]

.... Several of us have the CMS-50D which is available from Supplier 19.

I believe Paula means the CMS-50D Plus, which is the non-medical device. The CMS-50D is the FDA approved item, which is more expensive. I think.

Duh. Yeah, that's what I meant. I was testing your reflexes. Yeah, that's what I was doing.

I found this information in a pulse oximeter supplier's website that explains the differences in the Contec CMS-50D Plus and the models below (CMS-50D) and above it (CMS-50E) in Contec's product line.

What's the key difference between the CMS-50D Plus and the next model up?
The CMS-50D Plus is the most advanced pulse oximeter available for under $100. The primary difference between the CMS-50D Plus and the 50E model is that the 50E pulseox has a rechargeable lithium battery (compared to the regular alkaline batteries used in the 50D Plus). Also, the CMS-50E allows the user to enter the time for recording purposes. The CMS-50D Plus does not allow the user to enter the time of day.

What's the key difference between the CMS50D Plus and the next model down?
The CMS50D Plus features built-in flash memory, USB port and software. The next model down is the 50D which does not record data and which does not support computer connectivity.

So, for overnight use, you want the CMS-50D Plus (or CMS-50E) since they record data, whereas the CMS-50D does not (it is actually a cheaper model).

[SuperSleeper]

And see, if this fact is to be so important--and I agree it is--why should/would the machines record it differently? Users would not tolerate a glucose meter that is off. That company would be out of business. Why do we tolerate machines that record such an allegedly important number as the AHI differently? If one machine says my AHI is below 5 and another says it is above 5, which is right? Both? Neither?

Actually, glucose meters aren't all that accurate either, even when you use the same strips and meter, let alone when you switch between brands. As far as official standards, meter readings can be off +/- 20% and still be acceptable. From this article:

Blood glucose meters must meet accuracy standards set by the International Organization for Standardization (ISO). According to ISO 15197 Blood glucose meters must provide results that are within 20% of a laboratory standard 95% of the time (for concentrations about 75 mg/dL, absolute levels are used for lower concentrations).

That's a huge difference. That means that my meter can record my blood sugars anywhere between 80-120, roughly and it's just fine. I've tested my sugar levels using nearly the same spot on my finger or thumb using the same strips and the same meter, 2-3 times, and each time, I can get a substantially different reading: something like 120, 134 and 105, all tested at roughly the same point on my hand, using the same strips and meter within 3 minutes of each other.

For Sleep Apnea and trend analysis with the AHI, you don't need to have a completely and perfectly calibrated system of measure. All you really need is consistency within the algorithm of the CPAP you're using. That's good enough to spot trends. This is why we often tell patients not to go by the readings of a single night's AHI, or even two nights. We really need to come up with a larger statistical sample (time-wise) to get a more accurate trend. I won't make a pressure change unless I've been monitoring my AHI at a specific pressure for at least a week.

I actually think that there's not a huge difference in comparable readings of AHI amongst the different data-reporting CPAP machines. Maybe a point or two overall - but that doesn't really matter. Patients utilize the AHI readings of their one machine, generally, and when testing pressure and other settings, don't generally switch to another brand of CPAP. Thus, over time, using the same machine, they can actually get a pretty accurate representation of their own AHI trending data... at least enough to make a reasonable decision regarding whether or not a specific CPAP setting change has helped or hindered their therapy.

If one machine says my AHI is below 5 and another says it is above 5, which is right? Both? Neither?

It doesn't really matter. It's not an issue of "right" or "wrong"-- There is no sure-fire method for determining whether or not AHI is being reported correctly or not in a CPAP machine. Even if you were to compare (if possible) 2 overnight studies in a sleep lab with a full PSG (sleep test) in a fully-equipped clinical sleep lab, you'd get differences in the AHI numbers for the same patient. Labs don't have identical equipment all the time, so there's going to be variations.

The important thing for using AHI as a trend indicator is that the machine's algorithm and associated firmware be relatively consistent from night to night, which it generally is. It doesn't have to be calibrated exactly to the sleep lab's standards or even to other CPAP machines.

I'm over-simplifying in this example, I realize, but let's say, for instance, my actual treated AHI levels (on CPAP at 12 cmH2O), according to a fully functional sleep lab was 4.0. Now, I go home and put on my CPAP machine and it records my AHI as 3.5. Which is right? Doesn't matter. My home CPAP machine has now become my baseline standard.

I know that at 12 cmH2O of pressure my CPAP machine is off a little bit in reporting my AHI (by 0.5 AHI, roughly). I don't really care, because I'm going to take that 3.5 level and mess around with my pressure settings to try to lower that 3.5 AHI that my CPAP is reporting. I couldn't care less about the slight difference between my machine's AHI of 3.5 and the sleep lab's at 4.0. What I'm going to do is look at my machine's trends from now on. The sleep lab's data is meaningless to me for trend analysis, because I can't go in for a sleep test every week to check AHI levels. I use my home CPAP for that now. If increasing pressure to 13 gets my machine's AHI down to 2.5 AHI on average over a week, my machine's readings have proven effective at giving me enough trend data to make a logical pressure change. Doesn't matter that the actual AHI numbers disagree by a small margin with the sleep lab's numbers.

Hope that made some sort of sense. I get long-winded, I know.

[Sleepster]

And see, if this fact is to be so important--and I agree it is--why should/would the machines record it differently? Users would not tolerate a glucose meter that is off.

Glucose level is a single measurement. AHI is an index based on several factors, so it's a more complicated calculation. Hypopneas are probably the most complicated part of the AHI calculation. Many times I look at my hypopneas and wonder why they count in my AHI. I'm still breathing, so oxygen is still getting to my lungs. It's just that for 10-15 seconds my breathing has become a bit irregular. What's the logic used to score that as an event that contributes to my AHI?

[JumpStart]

Glucose may not be identical to AHI, but, in the overall picture, they are very similar. The complexity or lack thereof for determining AHI is irrelevant in determining trend, as SS noted, when the same algorithm and same sensors are used consistently. And, as pointed out, most users do keep the same machine for a significant period of time, thus everything remains the same. Were I to use a different glucose meter and a different manufacturer's test strip, everytime I needed a test, and chose a different body location for each sample, I could never establish an effective consistency, if that were my purpose in those spot checks. And even with consistentcy by the user, those glucose results are deemed acceptable over a very wide range. Fortunately, in glucose issues, we have A1c, which allows a blood sample to determine a (generally accepted) average covering a period of months, with spot checks by users serving a different specific purpose. A succession of A1c results can establish a trend. And we can determine the same effective trend via ResScan or SleepyHead from our machines (although not overnight). It is also irrelevant how any particular machine/brand counts apneas, hypopneas, and/or ca events. Just that they do so, and do so consistently.

[archangle]

Checking your apnea with AHI is a bit like checking your heart health by checking your blood pressure. If it's out of the normal range, or if it changes significantly, you better get it checked out.

You'd be foolish to only use BP testing if you have heart problems, but it's a good thing to check often.

Unfortunately, a lot of the medical communinity has their head up their * and don't bother to use the very useful additional tracking information available essentially for free with modern CPAP machines, and just look at AHI.

[CHanlon]

Glucose may not be identical to AHI, but, in the overall picture, they are very similar.

Just to work the glucose analogy a bit more... In that particular case, changing the test equipment can lead to serious errors an confusion.

I live in Canada, and while I don't have diabetes, I'm aware of a few people who do, both in Canada and in the US. So I know this little tidbit - Canada and the US use different units of measure for glucose. A Canadian that doesn't know any better using a machine they got in the US is gonna die...

In Canada, the target range for blood glucose is 4 - 6 (mmol/L) and in the US the same target range is 70 - 130 (mg/dl).

I point that out just to say that at least AHI is consistent. :-p