How to Read Your CPAP Data in OSCAR: A Beginner's Guide

Short version

OSCAR is a free program that turns the data your CPAP machine quietly records every night into graphs you can actually look at. You import data from your machine’s SD card (or sometimes directly from the device), then check three main things: your AHI, your leak rate, and any flow limitation or flag patterns. None of this replaces your sleep clinician, but it can help you understand what’s happening between appointments and ask better questions when you get there.

What OSCAR is, and what it isn’t

OSCAR (Open Source CPAP Analysis Reporter) is free software, maintained by volunteers, that reads the raw data files most CPAP and BiPAP machines store internally. You can download it from sleepfiles.com/OSCAR , where it’s available for Windows, Mac, and Linux.

It is not made by any CPAP manufacturer, and it’s not a diagnostic tool. It shows you the same underlying numbers your machine already calculates — it just displays them in a lot more detail than the small screen on your device or the companion phone app. Reading the graphs is a bit like checking your own blood pressure at home: useful information, but interpretation is still a conversation for your doctor.

I started using it a few months into therapy, mostly because the app that came with my machine gave me a single number each morning and nothing else. OSCAR was the first time I actually saw the shape of my night.

Getting your data into OSCAR

Step 1: Find your SD card. Most modern CPAP and APAP machines (ResMed AirSense, Philips Respironics DreamStation, and others) store detailed data on a removable SD card inside the unit. Check your device’s user manual for the card’s exact location — it’s usually behind a small panel on the bottom or back.

Step 2: Copy the data to your computer. Insert the card into a card reader on your computer. Don’t rename or reorganize the folders — OSCAR expects the manufacturer’s original file structure.

Step 3: Import into OSCAR. Open OSCAR, choose “Import,” and point it to the SD card or the folder where you copied the files. The first import can take a few minutes if you have months of data; after that, imports are quick since OSCAR only pulls in new nights.

Step 4: Put the card back. Your machine needs the card to keep recording, so don’t leave it out of the device for extended periods.

A few practical notes:

• Some machines (like newer ResMed models paired with myAir) sync data to the cloud instead of, or in addition to, the SD card. OSCAR generally works from the local card data, not the cloud account.
• If OSCAR doesn’t recognize your files, check whether your machine model is supported — the OSCAR project site and its help wiki list current models — before assuming something’s wrong with your setup.
• Back up your OSCAR data folder occasionally. Losing months of sleep history to a hard drive crash is annoying.

The daily overview: what you’ll see first

After importing, OSCAR opens on a calendar-style view, with each day showing summary numbers. Click on a night and you’ll get a detailed graph broken into rows: usage, AHI events, pressure, leak rate, flow limitation, and sometimes SpO2 or pulse if you use a connected pulse oximeter.

It looks like a lot at first. Most people end up focusing on three things regularly.

AHI (Apnea-Hypopnea Index)

AHI is the number of apneas (breathing pauses) and hypopneas (partial reductions in airflow) per hour of use, averaged over the night. The American Academy of Sleep Medicine’s obstructive sleep apnea fact sheet describes AHI as a common measure of sleep apnea severity.

In OSCAR, the AHI graph shows individual events as small marks along a timeline, plus a running average. The severity categories in that AASM fact sheet are:

• Under 5: not classified as sleep apnea in adults
• 5–15: mild
• 15–30: moderate
• Over 30: severe

These thresholds come from diagnostic sleep study criteria, not from CPAP therapy data specifically, but many people watch their treated AHI as a rough proxy for how well therapy is controlling their events on a given night. A single high night isn’t necessarily meaningful — colds, alcohol, sleeping position, and mask problems can all bump the number temporarily. A pattern over several weeks is more informative than any one night.

If your AHI is trending consistently higher than what your clinician set as your target, that’s a reason to bring the printout or a screenshot to your next appointment — not a reason to change your pressure settings yourself.

Leak rate

Leak rate measures how much air is escaping around your mask seal instead of going into your airway. Every mask has some intentional venting built in, which is normal and accounted for by the machine. What OSCAR flags is unintentional leak — from a poor seal, a shifted mask, or an open mouth.

ResMed machines typically display leak in liters per minute, and a rule of thumb many CPAP users and DME staff go by is that readings above roughly 24 L/min are worth trying to bring down. Philips Respironics machines report leak differently, so if you switch machine brands, the numbers won’t be directly comparable — check your model’s documentation rather than assuming the same threshold applies.

In the OSCAR graph, leak looks like a shaded band along the bottom of the chart. Spikes usually correspond to specific moments — rolling over, talking in your sleep, or a mask that’s slipped. A leak line that sits high for hours suggests a fit problem worth addressing, since heavy leak can both disrupt your therapy and make the machine’s own AHI calculation less reliable.

If you’re chasing a leak that won’t settle down, it’s often a mask fit issue rather than a machine issue. Our guide on stopping mask leaks into your eyes works through the usual fit culprits — cushion wear, headgear tension, seating, sleep position — which apply to a leak in any direction, not just toward the eyes.

Flow limitation and flags

Flow limitation refers to a flattening in the normal breathing wave shape, which can indicate the airway is narrowing even though it hasn’t fully closed into an apnea. OSCAR displays this as its own graph line, and most machines also generate “flags” — abbreviated markers like RE (RERA), FL (flow limitation), or CA (central apnea) plotted along the events row.

You don’t need to memorize every flag type. What’s worth noticing is clustering: a string of flow limitation flags around the same time each night, or a recurring pattern tied to sleeping position, can be useful detail to describe to your clinician. Some clinicians use this kind of detail to consider whether a pressure adjustment or a different machine mode might help — but that’s their call to make with your full clinical picture, not something to infer from a graph alone.

Putting it together: reading a night at a glance

A reasonably good night, from a data standpoint, often looks like:

• AHI well under the threshold your clinician described as your target
• Leak line low and flat for most of the night, with only brief spikes
• Few or no clustered flags in the events row
• Usage hours that reflect roughly what you actually slept

A night that might be worth a closer look:

• AHI noticeably above your usual average
• Leak elevated for long stretches, especially if you also woke up with dry mouth or noticed the mask hissing
• A cluster of central apnea (CA) flags, which some studies note can sometimes emerge or increase after starting CPAP therapy in certain patients — a phenomenon researchers refer to as treatment-emergent central sleep apnea , and worth flagging to your clinician rather than interpreting on your own

It helps to look at data in stretches rather than single nights. OSCAR lets you view a week or month at a time, which smooths out one-off bad nights caused by things like a cold, travel, or an extra glass of wine.

Common beginner mix-ups

Confusing the machine’s AHI with your original sleep study AHI. Your diagnostic AHI (from a sleep study, often without treatment) and your treated AHI (from your CPAP, with treatment) measure different things and aren’t directly comparable in a “before and after” sense, even though people often talk about them that way.

Assuming high leak explains everything. Leak affects data quality and comfort, but it isn’t automatically the cause of every bad night. Sometimes a high-leak night still has a low AHI, and a low-leak night still has a rough one.

Chasing a “perfect” AHI of zero. An AHI under 5 isn’t classified as sleep apnea in the first place, so a few scattered events on a given night aren’t a failure of therapy. A more realistic goal than a flat line every night is a steady, meaningful reduction in events alongside how you actually feel during the day.

Changing pressure settings based on OSCAR alone. This is worth repeating: OSCAR shows you data, but adjusting pressure, minimum/maximum EPR, or ramp settings is a decision to make with your sleep clinician or DME provider, who can weigh your data against your symptoms, health history, and the original prescription.

When to talk to your doctor or DME

Bring your OSCAR charts (screenshots or the exported PDF report work well) to your clinician if you notice:

• A sustained rise in AHI over multiple weeks rather than an isolated bad night
• Leak rates that stay high most nights despite trying different mask fits
• Recurring central apnea flags, especially alongside symptoms like gasping awake or unrefreshing sleep
• Daytime sleepiness or other symptoms that don’t match what the data seems to show — sometimes the two disagree, and that mismatch itself is useful information for your provider

Most sleep clinics are used to patients showing up with OSCAR data these days, and many appreciate the detail — it can make a follow-up appointment more productive than trying to describe “how I’ve been sleeping” from memory. Just keep in mind that your clinician is looking at your data in the context of your original sleep study, your health history, and your reported symptoms, which is more than any graph can show on its own.

If you’re new to OSCAR and want a gentler starting point before diving into every graph, it’s fine to wait a few weeks before you start paying close attention to the numbers. Settling into therapy — getting used to wearing the mask, stringing together a few normal nights — comes first. For me, the data made a lot more sense once I had those normal nights to compare against.