# Were Those CPAP Events Real, or Was I Awake?

A CPAP counts apneas and hypopneas from airflow and does not report if you were asleep. What that means for your events and how to look at your night.

Source: https://cpapclarity.com/learn/cpap-events-while-awake
Published: 2026-11-02
Updated: 2026-11-02
Author: Brian C.

Informational, not medical advice. Talk to your sleep physician before changing your therapy, pressure, mask, or medications.

Your CPAP machine scores apneas and hypopneas from the airflow it measures, and its report does not say whether you were asleep at that moment, so some counted events may have happened while you were awake. The machine cannot tell you which ones. That is a limit of what the machine records, not a verdict on your night. Below is what the machine measures, why breathing around falling asleep and waking can be counted, where to look in your own data, and what to bring to a sleep physician.

> **Not medical advice.** Don't change pressure, mask, or medications based on what you see here without talking to your sleep physician first.

> **Where these numbers come from** These numbers are read from your own device's export. Oxygen drop counts either come straight from your oximeter's own session summary (only for raw session files whose summary has been checked against the app; the page says so) or are computed by CPAP Clarity from the recorded readings; computed counts can differ slightly from what your device's own app shows, and when they disagree, treat the device's app as the reference. The therapy score is CPAP Clarity's own summary and has no device equivalent.

## What the machine actually measures

The scoring rules the sleep field uses were written for a sleep study. In the American Academy of Sleep Medicine (AASM) rules for adults, an apnea is a drop in the airflow signal of at least 90% from the pre-event baseline lasting at least 10 seconds. A hypopnea is a drop of at least 30% lasting at least 10 seconds, together with either a drop in blood oxygen of at least 3% or an arousal. During a pressure titration in a sleep lab, the AASM recommends the CPAP device's own flow signal as the sensor for finding apneas and hypopneas.

The lab setting matters. A sleep study also records brain activity, so a technologist can see whether you were asleep and can see arousals. The oxygen and arousal parts of the hypopnea rule come from those extra sensors. A home machine working from airflow alone has no direct way to check them the same way, and the flow trace carries no label saying whether the person breathing was asleep.

There is one place where a machine does make a guess about sleep. ResMed's AirSense 10 user guide says that when Ramp Time is set to Auto, the device will detect when you have fallen asleep and then raise the pressure to the prescribed level. That guide does not say the guess is used to decide which events get counted, and the report you see does not mark any event as asleep or awake.

## Hours of use are not hours of sleep

The AASM defines the apnea-hypopnea index (AHI) as apneas plus hypopneas per hour of sleep. ResMed and Philips reports are built on hours of use instead. ResMed's myAir report guide defines usage as the length of time you receive therapy from your machine, and a 2022 study of Philips Respironics machines described the machine's event index in events per hour of use. Those definitions do not subtract time spent awake. If you lie awake under the mask for 40 minutes, those 40 minutes sit in your hours of use whether or not you slept. ResMed's guides define usage but, in the pages checked for this article, do not state the exact hours the AHI is divided by, so check your machine maker's documentation if the exact denominator matters to you.

If a machine's index is expressed per hour of use, the effect of awake time cuts in two directions at once. Events counted while awake add to the top of the fraction. Awake time under the mask adds to the bottom. From the report alone you cannot tell which effect is bigger on a given night, and it does not have to be the same from one night to the next. So this article does not claim that awake events make your AHI look higher or lower. It only says the machine's number is not built on time asleep the way the AASM's definition is.

## Why breathing near wake can be scored

Breathing is not steady in the minutes around falling asleep. In a 2005 study in the Journal of Applied Physiology, Thomson and colleagues followed eight healthy people through the drowsy period before sleep onset. Breath by breath, when brain activity on the EEG shifted from an alpha to a theta pattern (the drowsy state), the length of a breath went from about 5.2 seconds to about 13.0 seconds on average, and the pause after breathing out went from about 0.7 seconds to about 7.5 seconds. The authors conclude that the carbon dioxide level at which breathing pauses (the apneic threshold) sits at or just below the level present when awake. These were healthy volunteers on a ventilator in a lab, not CPAP users, so the numbers do not transfer to your machine. The longest pause in flow they measured averaged 7.5 seconds, short of the 10 seconds the apnea rule asks for. The study shows that breathing is unstable at sleep onset in healthy people; it does not show that a machine would score an event there.

Other everyday changes to the pattern of air moving through a mask can also show up in the flow trace, such as talking, a deep sigh, a cough or shifting position. The machine records that the flow changed, and it does not record why.

How well machine counts match hand scoring is a related question, and the answer is not simple. In a 2022 study in the Journal of Clinical Sleep Medicine, Ni and Thomas followed 179 adults with obstructive sleep apnea on Philips Respironics machines, whose untreated AHI had a median of 46 at their split-night sleep study. The machine's respiratory event index (a flow-based count, not the AHI) ran 10.72 events an hour higher than manual scoring in the first month (standard deviation 8.43), and the gap held steady for up to 12 months. Larger gaps were more common with male sex, a large leak, a higher arousal index at the titration study and a higher ratio of unstable to stable breathing. That study did not isolate wake as a cause, it covered one manufacturer's machines, and it does not say what your machine would do. It does show that, in that group, the machine's count and a trained scorer's count differed by about 10.7 events an hour on average.

## How to look at your own night on CPAP Clarity

CPAP Clarity shows the machine's count as the machine reports it. It does not re-score your night or remove events, and it cannot say whether you were awake for any of them. What it can do is show you when each event fell, so you can look for a pattern and take it to your provider.

- **The night ribbon.** On your night page, the card called "Your night, hour by hour" puts the night on one clock. A quiet night draws each event as a dot, lettered by type when your machine reports types; a busy night draws bars that each count the events in a short stretch. The left and right ends read the mask-on and mask-off times when your machine records them (otherwise the ends read as the start and end of the recording). If the machine split the night into sessions and recorded the mask times, the gap where the mask was off is shaded. Moving the pointer along the ribbon reads out the time, and also the stage and the oxygen level if you have a wearable or oximeter connected.
- **The event list under the ribbon.** Up to ten events are listed by time, and when you have a wearable each one carries the wearable's stage at that moment (for example "had you awake"). That is the most direct place to see where an event and a wearable's awake estimate fall together.
- **The timeline in Detailed Charts.** Open Detailed Charts and the timeline row shows every scored event as a dot on the same clock as your other charts. Its step buttons walk the night one event at a time, so you can read each event's time.
- **A wearable's stage band.** If you import a wearable such as a ring or watch, the ribbon can show its "Awake" periods beside your events. That band is the wearable's own estimate, built from signals like movement and heart rate, so an event under an awake stretch shows where two devices disagree, which is worth a note and does not settle anything.

Things worth noticing are events packed into the first minutes after mask on, events right around a mask-off gap or just before the mask comes off in the morning, and whether the same shape repeats over many nights or showed up once. Treat each of these as a question to bring to your provider and not as an answer. An event near the start of the night might have happened awake, or you might have fallen asleep quickly. The data on its own cannot say which.

## What this does and does not mean for your AHI

It does not mean your AHI is wrong. It does not mean it is right either. The number is the machine's count of flow events, reported over your hours of use, and it is the same kind of number your provider sees on the machine's report, so changes from your own earlier nights are the comparison your provider will usually look at. Reading it as a count of events during sleep, the way a sleep study's index is defined, goes a step past what the machine measured.

Please do not subtract events from your own report to make a new AHI. Which events to leave out would be a guess, and the hours of use would need the same adjustment. If your AHI and how you feel do not seem to match, the useful step is to take the pattern to someone who can order a test that does record sleep. Our guide to [feeling tired with a good AHI](https://cpapclarity.com/learn/good-ahi-but-still-tired) covers the other signals a CPAP cannot see, and [what AHI is](https://cpapclarity.com/learn/what-is-ahi) explains how the number is built. If most of your counted events are central apneas, [central versus obstructive apnea](https://cpapclarity.com/learn/central-vs-obstructive-apnea) explains what that split means, and [how to read the flow rate graph](https://cpapclarity.com/learn/how-to-read-cpap-flow-rate-graph) shows what a single event looks like in the raw trace.

## When to talk to your sleep physician

Raise it with your sleep physician or the provider who manages your therapy if any of these fit.

- You feel unrefreshed or sleepy in the day even though your AHI looks low.
- Your AHI looks high but you spend a long time awake in bed, or you notice your events pile up right after you put the mask on.
- You see the same clustering night after night, and you want to know whether it matters for you.
- You are not sure whether your machine's report is telling you what you think it is.

Your provider can decide whether anything about your settings or your testing should change. A sleep study that records brain activity is the test built to separate events during sleep from events during wake, and whether you need one is your physician's call. Please do not change pressure settings yourself on the strength of a pattern in your data.

[Import your card and look at your own night, free →](https://cpapclarity.com/)

## Common Questions

### Does my CPAP know when I am asleep?

Not in the way a sleep study does. The machine records airflow and pressure, and its report does not mark events as asleep or awake. ResMed's AirSense 10 user guide says its Auto ramp setting detects when you have fallen asleep and then raises the pressure, but that guide does not say the guess is used to decide which events are counted.

### Can events counted while I was awake change my AHI?

They can add to the count, and the awake time under the mask also adds to your hours of use, so the net effect on a given night is not something the report can show. The AASM defines AHI as events per hour of sleep, and machine reports are usually built on hours of use. Check your maker's documentation for its exact denominator.

### Is a 10-second pause at sleep onset a real apnea?

The AASM rule asks for a drop of at least 90% in the airflow signal lasting at least 10 seconds. In a 2005 lab study of eight healthy volunteers on a ventilator, breathing became unstable as brain activity shifted toward sleep, and the longest average pause in flow was 7.5 seconds. That study was not on CPAP and did not test the 10-second rule. Whether a gap on your machine is counted depends on its method, and your physician can judge what your pattern means.

### Do machine event counts match hand scoring?

Not always. In a 2022 study of 179 adults on Philips Respironics machines (median untreated AHI 46), the machine's respiratory event index averaged 10.72 events an hour higher than manual scoring in the first month, and the gap lasted up to 12 months. It was a respiratory event index, not the AHI, and the study did not isolate wake as a cause.

### Should I ignore the events at the start of my night?

No. Please do not remove or discount anything on your own. Events near the start of the night may have happened awake, or you may have fallen asleep quickly, and the report cannot show which. Note the pattern and bring it to your provider.

## Primary Sources

- Berry RB, Budhiraja R, Gottlieb DJ, et al. *Rules for scoring respiratory events in sleep: update of the 2007 AASM Manual for the Scoring of Sleep and Associated Events.* J Clin Sleep Med. 2012 Oct 15;8(5):597-619. The adult apnea rule (a drop of at least 90% for at least 10 seconds), the adult hypopnea rule (a drop of at least 30% for at least 10 seconds with a 3% oxygen desaturation or an arousal), the PAP device flow signal as the recommended sensor during titration, and AHI defined as events per hour of sleep. [PubMed 23066376](https://pubmed.ncbi.nlm.nih.gov/23066376/)
- Ni YN, Thomas RJ. *A longitudinal study of the accuracy of positive airway pressure therapy machine-detected apnea-hypopnea events.* J Clin Sleep Med. 2022 Apr 1;18(4):1121-1134. 179 adults with OSA (median baseline AHI 46.3) on Philips Respironics machines; machine-detected respiratory event index 10.72 events an hour higher than manual scoring in month one, stable to month 12. [PubMed 34886948](https://pubmed.ncbi.nlm.nih.gov/34886948/)
- Thomson S, Morrell MJ, Cordingley JJ, Semple SJ. *Ventilation is unstable during drowsiness before sleep onset.* J Appl Physiol. 2005 Nov;99(5):2036-44. Eight healthy subjects; breath duration rose from 5.2 to 13.0 seconds and the expiratory pause from 0.7 to 7.5 seconds at the alpha-to-theta EEG transition. [PubMed 16024516](https://pubmed.ncbi.nlm.nih.gov/16024516/)
- ResMed. *AirSense 10 user guide* (the Ramp Time Auto description). [ResMed AirSense 10 user guide (PDF)](https://document.resmed.com/documents/products/machine/airsense-series/user-guide/airsense-10-device-with-humidifier_user-guide_amer_eng.pdf)
- ResMed. *myAir report guide* (definition of usage). [ResMed myAir report guide (PDF)](https://ap.resmed.com/hubfs/Technical%20Service/myAir%20Report%20Guide.pdf)