Most CPAP machines store detailed therapy data on a standard SD card, and downloading that data to your computer takes only a few minutes once you know where to look. The process is straightforward: remove the card from the machine, insert it into a card reader, and open the files with free software designed to interpret them. The real value, though, comes from what you do with the data once you have it, since the detailed breath-by-breath records on that card contain far more information than the summary screen on your machine ever shows.
Finding the SD Card in Your Machine
Every major CPAP brand uses an SD card to log therapy data, but the card slot lives in a slightly different spot depending on the manufacturer. On ResMed machines like the AirSense 10 and AirSense 11, the SD card slot is on the left side of the device, behind a small flip-open door. On older Philips Respironics DreamStation units, the card slot sits on the back of the machine near the power connection. The newer DreamStation 2 uses a micro SD card tucked behind a rubber flap on the side. Most machines ship with a card already installed, so if you have never looked for one, it is probably in there right now.
Before you pull the card out, turn the machine off. Removing the card while the device is writing data can corrupt files. Once the machine is powered down, press the card gently and it will click and spring out slightly, just like a phone’s SIM tray. If the card feels stuck, check whether a small latch or door needs to be opened first. Do not force it.
Reading the Card on Your Computer
Standard SD cards fit directly into the built-in card reader on many laptops. If your computer does not have a slot, a USB SD card reader costs a few dollars and plugs into any USB port. For machines that use micro SD cards, you will need either a micro SD reader or a micro-to-full-size adapter, which usually comes packaged with the card.
When you insert the card, your computer should recognize it as a removable drive, just like a thumb drive. You can open it in your file browser, and you will see folders with names that vary by manufacturer. ResMed machines create folders labeled by date, with files ending in extensions like .edf and .crc. Philips machines store data in folders with less intuitive names. The raw files are not meant to be opened in a spreadsheet or text editor. They are encoded in proprietary or semi-proprietary formats, which is why you need dedicated software to make sense of them.
Free Software That Reads CPAP Data
The most widely used tool in the CPAP community is OSCAR (Open Source CPAP Analysis Reporter). It is free, runs on Windows, Mac, and Linux, and supports machines from ResMed, Philips, Fisher & Paykel, and several other brands. To use it, download OSCAR from its official site, install it, insert your SD card, and point the software to the card’s drive letter. OSCAR imports the data automatically and presents it in charts and graphs that are far more detailed than anything on your machine’s screen or your provider’s portal.
OSCAR shows you nightly summaries and breath-by-breath waveforms, including your apnea-hypopnea index (AHI), mask leak rates, pressure levels throughout the night, tidal volume, and respiratory rate. You can zoom into any five-minute window of your sleep session and see exactly what your breathing looked like, whether the machine ramped up pressure in response to an event, and how your leak levels fluctuated. For people who want to understand their therapy rather than just comply with it, this level of detail is a game-changer.
Manufacturers also offer their own apps and platforms. ResMed’s myAir app pulls data wirelessly from newer machines and gives you a simplified daily score. Philips offered DreamMapper for similar purposes before its recall-related disruptions. These apps are convenient but deliberately simplified. They show you a nightly score and broad trends, not the granular event-by-event data that OSCAR provides. If you want to troubleshoot a specific problem, like why you wake up at 3 a.m. feeling like your mask is blasting air, the manufacturer app will not get you there. OSCAR will.
What the Data Actually Contains
The SD card stores more information than most users realize. Here is what you will find once you import it into software:
- AHI breakdown: Your machine records obstructive apneas, central apneas, and hypopneas separately, and timestamps each event. The overall AHI number on your machine’s screen is a nightly average, but the detailed log lets you see whether events cluster at certain times, such as during REM sleep or after you roll onto your back.
- Pressure data: If you use an auto-adjusting (APAP) machine, the card logs the pressure delivered at every moment throughout the night. You can see your median pressure, your 95th-percentile pressure, and how often the machine hit its maximum setting. This is useful for spotting whether your pressure range needs adjustment.
- Leak rate: The machine tracks how much air escapes from your mask seal. Leak data includes both intentional leak (the air vented through your mask’s exhaust ports by design) and unintentional leak (air escaping around the edges because the mask shifted or your mouth opened). Research on leak metrics distinguishes between these by subtracting the mask-specific intentional leak from the device-reported average, isolating the unintentional portion that actually affects therapy quality.1PubMed Central. What’s leak got to do with it? Association of mask leak and positive airway pressure adherence from the homepap study
- Usage hours: The card logs exactly when you turned the machine on and off, and how many hours of effective therapy you received each night. This is the number your insurance company and sleep clinic care about most.
- Flow waveforms: The most granular layer of data. OSCAR can display the actual airflow signal, showing each breath you took. This lets you or your clinician spot patterns like flow limitation (partial obstruction that does not quite qualify as an apnea but still disrupts sleep) or periodic breathing patterns.
Why Downloading Your Own Data Matters
There is a practical reason beyond curiosity. CPAP adherence, defined by most insurers as using the machine for at least four hours per night on at least 70 percent of nights, determines whether your insurance continues to cover your equipment. Having your own copy of the data means you are not relying solely on your provider’s portal to verify your numbers. If there is a dispute about compliance, your SD card is a primary record.
Beyond insurance, engaging with your own data tends to improve how consistently you use the machine. A retrospective study of patients using an active engagement platform found that those who interacted with their therapy data achieved adherence at a rate of about 87 percent, compared with roughly 70 percent among patients receiving usual care, and averaged about an hour more of nightly use.2PubMed Central. Patient Engagement Using New Technology to Improve Adherence to Positive Airway Pressure Therapy: A Retrospective Analysis Patients who were struggling with adherence saw even larger gains. Downloading and reviewing your data is not the same as using a manufacturer’s engagement platform, but the underlying principle holds: people who see what their therapy is doing tend to stick with it.
Common Problems When Downloading
A few issues come up repeatedly. The most common is that the computer does not recognize the SD card at all. This usually means the card reader is not working, the card’s contacts are dirty, or the card has developed a fault. Try a different reader first. If that does not help, gently clean the card’s gold contacts with a dry cloth and reinsert it. SD cards do fail over time, especially the low-cost ones that ship with CPAP machines. Replacing a worn-out card every year or two is reasonable.
Another frequent problem is OSCAR refusing to import the data. This happens when the software version does not yet support your specific machine model, which is common with very new devices. OSCAR is maintained by volunteers and updated regularly, so checking for the latest version often solves it. If your machine is a brand-new release, you may need to wait a few weeks for the developers to add support.
Some users accidentally format the SD card when inserting it into their computer, especially if a prompt pops up saying the card needs to be formatted before use. Do not format it. That erases all your data. The card uses a file system the machine understands, and your computer can read it without reformatting. If you accidentally formatted the card, the data is gone unless you use a file-recovery tool quickly, and even then results vary. The safest habit is to copy the entire contents of the card to a folder on your computer before doing anything else. That way you always have a backup.
Occasionally, the card fills up. Most CPAP SD cards are 4 to 8 gigabytes, which holds roughly a year or more of data depending on the machine and how much detail it records. When the card is full, some machines overwrite old data automatically, while others stop recording. If your card is getting full, copy the data off and either continue using the same card (the machine will start fresh) or swap in a new one. Cards up to 32 gigabytes work in most machines, though going larger than that can cause compatibility issues with older models.
Making Sense of Leak Data
Leak is the metric that confuses people most, partly because different machines report it differently. ResMed machines report leak in liters per minute and include intentional leak in the total number. Philips machines historically reported only unintentional leak. When you see a leak number in OSCAR, check which definition the software is using for your particular machine, because a “leak” of 24 liters per minute on a ResMed might be perfectly normal (most of it is intentional exhaust), while the same number on a Philips machine would indicate a serious mask problem.
High unintentional leak degrades therapy because the machine cannot maintain the prescribed pressure if air is escaping. It compensates by increasing flow, which can dry out your airway, cause aerophagia (swallowing air), and make the machine louder. If your data shows leak spikes at consistent times each night, you are probably shifting position in your sleep and breaking the mask seal. If the leak ramps up gradually over several hours, the mask straps may be loosening as the material warms, or your jaw may be dropping open during deeper sleep stages. A chin strap or a full-face mask can address the jaw issue. Retightening the straps is not always the answer, since an overtightened mask often leaks worse because the seal deforms under pressure.
What Your Clinician Sees Versus What You Can See
Your sleep clinic typically gets a summary report, either wirelessly through the machine’s cellular modem or when you bring in your SD card at an appointment. That report shows nightly AHI, average usage hours, and leak percentages, often compressed into a one-page overview. It is designed for a clinician to glance at during a brief follow-up visit and decide whether therapy is working.
The SD card data is the raw layer beneath that summary. A clinician reviewing the full data in OSCAR or the manufacturer’s clinical software (like ResMed’s AirView) can see patterns that the summary hides. For example, your overall AHI might be two events per hour, which looks great, but the detailed data might show a cluster of central apneas in the first hour after you fall asleep, suggesting a different issue than obstructive sleep apnea. Or your average leak might be acceptable, but the data shows 20-minute stretches of high leak followed by the machine ramping to maximum pressure in a futile attempt to compensate, which wakes you up. These patterns matter, and they are invisible in summaries.
Researchers have even explored using CPAP airflow data to detect conditions beyond sleep apnea. A systematic review examined whether Cheyne-Stokes respiration patterns, captured in the airflow signals recorded by CPAP machines, could serve as a marker for heart failure in patients with obstructive sleep apnea.3PubMed Central. Cheyne-Stokes respiration detected via CPAP devices as a digital biomarker for heart failure in obstructive sleep apnoea: systematic review The idea is that the breathing data your machine quietly records every night could flag a serious cardiac condition before it causes obvious symptoms. That is still in the research phase, but it hints at how much clinical information sits on that small card.
Backing Up and Organizing Your Data Long-Term
OSCAR stores imported data in its own database on your computer, so once you have imported from the card, the data lives in two places: the card and your hard drive. Keeping both copies is wise. Some users import weekly; others do it monthly. The main thing is not to let years accumulate on the card without backing up, because SD cards are physically fragile and do fail without warning.
If you switch machines, your old data remains viewable in OSCAR. The software keeps separate profiles for each device, so you can compare how your therapy looked on your old machine versus your new one. This is especially useful if you switch from a fixed-pressure CPAP to an auto-adjusting model and want to see whether the new machine’s algorithm is handling your events better.
For people who travel with their CPAP, the SD card also functions as a portable record. If you visit a sleep specialist in another city or country, handing them the card gives them your complete therapy history. It is more reliable than trying to get records transferred between clinics, and it contains more detail than any summary report.
When Auto-Adjusting Machines Get It Wrong
One of the more useful things you can spot by downloading your data is whether your auto-adjusting machine is responding appropriately to breathing events. These machines use algorithms to detect airflow limitation and increase pressure accordingly, but the algorithms are not flawless. Bench testing of several auto-CPAP devices found meaningful differences in how they detected flow limitation and adjusted pressure, with some devices increasing pressure in situations that did not warrant it.4Chest. Bench evaluation of flow limitation detection by automated continuous positive airway pressure devices One device in that evaluation showed autotriggering, raising pressure in response to artifacts rather than genuine breathing events.
In practical terms, this means your machine might be running at higher pressures than you need, which can increase leak, cause discomfort, and make the machine harder to exhale against. If your OSCAR data shows the machine frequently hitting the top of its pressure range without a corresponding spike in respiratory events, that is a conversation worth having with your sleep doctor. They can narrow the pressure range, set a fixed pressure, or switch you to a different device whose algorithm better matches your breathing pattern. Without the data, you would just know that the machine feels uncomfortable some nights. With the data, you can see exactly why and point your clinician to the specific timestamps.