HI vs AI

I have been using a brand new DeVilbiss CPAP machine for the last few weeks.. I have all the extra attachments (software, SD Card, etc.) so I've been monitoring my sleep... I noticed my Aprea Index is quite low (.8 to 3/hr) but my Hypopnea Index (HI) is usually quite higher... 4.5 to 10 per hour.... pushing my total AHI up accordingly. My Pressure limit is set at 7 to 14, although it hovers around 8, and rarely goes as high as 13 or 14.... Mask leak is good... but can't explain why the high HI nd how I can lower that. Any ideas? BTW, my 2 flex settings are set at 3.... no delay at startup and the heater at around 3. I am also using a new mas: the ALOHA which I like very much.

I am not familiar with your machine, but, I was using a Resmed S8 machine that definitely overstated my HI. My AHI was always 5.0+. Once I upgraded to a Resmed S9, my AHI consistantly dropped to 1.5-2.5 with the same settings and mask that I used with my S8.

LSAT, I have a Resmed S8 as well and my HI is consistently higher than my AI (which is sometimes 0). My results for the last 7 days were AHI 3.3 with AI .2 and HI 3.1. How do you know that the S8 is overstating the HI? Is it not possible that your S9 is actually providing more effective therapy and that's why the HI is lower than with the S8? Or is it a known issue with the S8 that it overstates HI?

It is a known issue with the S8. One of our experts here always recommends that you automatically divide the HI in half.

I have worked with several people that had higher than usual Hyponea counts with the Devilibiss machines when the OA count is pretty much non existent. I haven't been able to figure out why. It doesn't happen to everyone though. Just like the S8 high hyponea count didn't happen to everyone. I don't know if it is related to the different algorithm and slightly different criteria or it's just a coincidence. Mostly what we have done is adjust the pressure upwards just a little to bring down the hyponea count just a little (like we would any other obstructive event) and as long as the person was sleeping well and feeling decent we just didn't worry that much about the hyponea count if it was within acceptable limits.

Edit for clarification. If using apap mode...a little more minimum pressure. Not maximum. If using cpap mode a little cpap pressure.

Thank you for the comments and replies... it's odd that the machine would read things wrong...
It baffles me how increasing the limits may have an impact especially since the machine is not going anywhere near the maximums.. so increasing it even higher may not be the answer. You'd think all machines work pretty much the same in calculating a 'standard' measurement like HI.

I'd like to better understand the difference between AI and HI.. can anyone explain?

DHT wrote:Thank you for the comments and replies... it's odd that the machine would read things wrong...
It baffles me how increasing the limits may have an impact especially since the machine is not going anywhere near the maximums.. so increasing it even higher may not be the answer.

I think the suggestion is that you should increase the minimum setting a bit so that there is always a bit more pressure when you're at the low end of the auto range.

You'd think all machines work pretty much the same in calculating a 'standard' measurement like HI.

Well, the AASM itself has two different ways of scoring hypopneas on an in-lab sleep test

As far as machine scoring of H's is concerned though: Differences can creep in to the various algorithms based on such things as

• How many minutes does the machine use to calculate the base line values for the wave flow?
• Exactly how is the base line calculated after the sampling window is chosen? Median? Average? 90% of the inhalations?
• Exactly how much decrease in airflow has to happen before an H is scored? At least 30%? 40%? 50%?
• Where is the line drawn between scoring an H versus an OA? An 80% drop in airflow? A 90% drop?

I'd like to better understand the difference between AI and HI.. can anyone explain?

An apnea is when the airflow actually stops for 10 seconds or more when you are asleep. A hypopnea is an OA wannabe: The airflow drops substantially for at least 10 seconds, but does not actually stop; like an apnea, a hypopnea must happen when you are asleep to be scored on an in-lab sleep test.

So on an in-lab sleep test:
AI = (# of apneas scored)/(total time the EEG says you are asleep)
HI = (# of hypopneas scored)/(total time the EEG says you are asleep)

On our home machines there is no EEG data and the machine has no idea if you are asleep or awake when you are using it. So on our machines:
Machine AI = (# of apneas scored)/(total time the machine was running)
Machine HI = (# of hypopneas scored)/total time the machine was running)

DHT wrote:Thank you for the comments and replies... it's odd that the machine would read things wrong...
It baffles me how increasing the limits may have an impact especially since the machine is not going anywhere near the maximums.. so increasing it even higher may not be the answer. You'd think all machines work pretty much the same in calculating a 'standard' measurement like HI.

I'd like to better understand the difference between AI and HI.. can anyone explain?

AI is when your breathing stops...HI is when your air flow is reduced.
As to your comment "its odd that the machine would read things wrong". It doesn't...It's the algorithm that is written into its logic. Each manufacterer
puts a program in their machine to measure AI HI CA etc.

One thing is that what you see as events on your CPAP machine (OA's, CA's, and H's) are the ones that your machine could not treat, you actually experienced them. So, assuming all was equal in the detection area (a major assumption, I realize), if you were getting better therapy, you'd see less of any or all of them, since more of them would be taken care of by the machine.