The math of calculating AHI

[PaulaO2]

We all know how to reach the AHI number and even how to back up and determine a guestimate of the number of events.

I was just researching something for a new member and came across an overly complicated way to calculate AHI:

To determine AHI, add the total number of apnea events, plus hypopnea events and divide by the total number of minutes of actual sleep time, then multiply by 60.

Example:

Apnea + Hypopnea divided by actual sleep time, then multiply by 60
200 apneas, 200 Hypopneas (400 Total Events)
420 Minutes Actual Sleep Time (7 hours x 60)
Divide 400 by 420 = .95 x 60 = 57 AHI (Severe OSA)

At first it made no sense. The info was on the website of a sleep lab. Going on the assumption that they track minutes, not hours, it makes slightly more sense, but only a little. They reach the number of minutes by taking the hours and multiplying by 60. Then putting that 60 right back in again. I can see the usefulness if, say, someone slept 6 hrs and 43 minutes. It would indeed help with that.

So I tried it out.

Let's say I had 200 apneas, 200 hypopneas for a total of 400 events.
I slept for 6hrs and 43 minutes. That would be 6x60=360+43=403.
400 events divided by 403 minutes = .992 then x 60 = AHI of 59.55

Now, let's round up. 400 / 7 = 57.14

Really, does it matter to be that freakin' exact? Anything over 30 is considered severe (I keep saying they need to revise that) so, does it matter to make a difference 2.41?

[JJJ]

Really, does it matter to be that freakin' exact? Anything over 30 is considered severe (I keep saying they need to revise that) so, does it matter to make a difference 2.41?

Probably not, although the severity of the patient's condition determines the initial prescription for the pressure settings. Bear in mind that the vast majority of people with sleep apnea just do what their doc says to do and never question it or try to learn how to tweak their settings to maximize the benefit of the therapy. Therefore, accuracy is of some benefit. And since it costs nothing to be more accurate, why not?

I constantly complain to myself that the data provided by my Autoset S9 is not accurate. There is only so much the machine can do. I'm sure there are events recorded as apneas or hypopneas that were really not apneas or hypopneas, and vice-versa. The number is probably small enough not to worry about. But I still strive for as much precision as I can get.

And, in spite of knowing that the AHIs reported by my machine may not be perfectly accurate, I still strive for the elusive zero AHI night. I came the closest ever last night when SleepyHead told me I had 0.09. I was thrilled, in spite of the fact that I felt terrible because I had a miserable night's sleep due to missing my bedtime. (Went to a Wagner opera and didn't get home until 1 am.)

You're right, it doesn't matter much. But in science more precision is always better.

[archangle]

The person who came up with that was probably not very good at math. Or was simply trying to make it "simple" for the people filling out the forms.

They may have also been worried about people dealing with difficult concepts like "200 minutes is 3.333333333 hours" vs. "200 minutes is 3 hours 20 minutes."

Decimals and hours:minuteseconds is scary. Then you put them there decimals together with ciphering, it's terrifying to Jethro.

It's still total events divided by hours.

It would actually make the most sense to do it as events times 60 then divided by total minutes. There are fewer chances for transcription errors if you don't have those pesky decimal numbers.

[Shastzi]

In one isolated case it might not make that much difference but if each measurement is off by 5% and there are thousands of
events across a time series AND you start pulling statistics from that, the statistics will have a very unhealthy skew to them.

ie: garbage in = more garbage out.

[Schnauzers 5]

Does my card do it for me? My doc seemed to have the info from the little card they took out of my c-pap. My card also shows I am totally compliant every day/night.

[JJJ]

Does my card do it for me? My doc seemed to have the info from the little card they took out of my c-pap. My card also shows I am totally compliant every day/night.

If you'd install SleepyHead you could view the data yourself. You'll be amazed at how you can improve your therapy.

[Schnauzers 5]

Does my card do it for me? My doc seemed to have the info from the little card they took out of my c-pap. My card also shows I am totally compliant every day/night.

If you'd install SleepyHead you could view the data yourself. You'll be amazed at how you can improve your therapy.

Ya know, I am not at all computer savvy and have no clue what I would be seeing to interpret. I do know my AHI is always under 5 and low and I have to think I am doing as well as possible. when i get a new doc I might go for a new evaluation sleep over but would prefer sleeping at home for a home study if they are ok. and have good info. I imagine I would sleep better in my own bed. It is allll daunting I must say. I have peeked at my card and have no clue what I am seeing or my records either. My oxygen levels ate good.

[Sleepster]

Ya know, I am not at all computer savvy and have no clue what I would be seeing to interpret.

If you can navigate through Apnea Board, reply to and quote from messages, you can evaluate your data using SleepyHead or EncoreBasic.

I do know my AHI is always under 5 and low and I have to think I am doing as well as possible.

As long as your leaks are under control you're probably right. You can look at your machine's on-screen menu and see the percent time in large leak. It should be zero, or close to it.

[Sleepster]

Really, does it matter to be that freakin' exact?

Nope. But in some cases it does make the math easier.

Suppose you had 16 events in 160 minutes.

[jjtisch3]

I can see total events, the time, oxygen levels and so forth on my wrist oximeter printout. I know the hypopnea events have to be added to the total events to get started to find my AHI. But, I cannot find anything that says hypopneas on my printout. 
Thank You
John