What 1,493 CPAP Users Taught Us About Mask Fitting
Over six weeks we asked every customer three questions about their CPAP mask experience. The results confirmed some things we expected — and directly contradicted advice that's been repeated for years. Here are the six findings that changed how we think about fitting.


The questions were simple:
- How many masks did you try before finding the right one?
- What made your previous masks fail?
- What's still waking you up at night?
We offered $5 in rewards points for completing all three. We did not prompt or guide the answers beyond the multiple-choice options provided.
What came back was the clearest picture of real-world CPAP mask experience we've seen — from patients at the moment of purchase, not in a clinical trial, not filtered through a provider's interpretation.
Finding the right mask is normal work — not a sign of failure
The single most common CPAP mask experience in our data is the Standard Search: trying 2–3 masks before finding one that works.


47% of patients fall between trying 2-3 masks. Another 11% went through 4 or more. Only 28% found their ideal mask first try.
What it Means
Patients who are on their second or third mask are at the center of the distribution, not the outlier. Setting the expectation that 2–3 masks is normal reduces anxiety and improves therapy adherence during the search phase.
Changing masks usually doesn't fix a leak — the cause travels with the patient
Leaking was named as the primary deal-breaker by 34% of all patients. Among the 507 who cited leaking, 63% still report a Ghost Leak on their new mask. The Ghost Leak appears at 39–62% across every patient segment — including 46% of patients whose previous mask was described as perfect.
The three most common Ghost Leak causes — wrong cushion size, over-tightened straps, worn cushion — are independent of the mask model. Changing the mask doesn't address any of them.
What it Means
A protocol that diagnoses the leak source before recommending a new mask would prevent unnecessary purchases. Most Ghost Leaks resolve with a cushion swap ($15–35) or a free sizing check, not a new frame that will cost you more.
Sleeping position doubles the difficulty of finding a working mask
Patients whose deal-breaker was sleeping position — masks that shifted when they turned — are more than twice as likely to still be searching (26.4%) vs. all others (12.9%). Their first-try success rate is 9.5%, compared to 30% for all other patients.
What it Means
Sleeping position should be the first filter in any mask selection process. Side sleepers need a different starting shortlist — and need their hose management addressed as a separate problem from the seal.
"Finding the right mask" doesn't end overnight disruption
The 273 patients who found their ideal mask on the first try with zero complaints represent the best-case CPAP experience in our data. Among them, 46% still wake to a Ghost Leak and 21% report a Hose Tug as their primary overnight nuisance. A perfect mask fit does not eliminate overnight disruption.
What it Means
Post-fitting care should explicitly address tube management and cushion maintenance schedules — not just congratulate the patient on finding their mask. The disruptions in Layers 2 and 3 are predictable and preventable.
Bridge pain and leaking often coexist in a fixable feedback loop
Among the 361 patients whose deal-breaker was comfort failure (marks, bridge pain), 25% still report a Ghost Leak as their top overnight nuisance — simultaneously. This reflects a specific loop: patients tighten straps to stop a leak → cushion compresses → bridge pain develops → loosening reopens the leak. Neither problem resolves through strap adjustment alone.
What it Means
Patients presenting with both bridge pain and active leaking should be assessed for cushion sizing before any strap adjustment recommendation. The "tighten spiral" is self-perpetuating until the root cause is addressed.
9.3% of patients report their biggest problem is their partner's comfort
What it Means
Patients presenting with both bridge pain and active leaking should be assessed for cushion sizing before any strap adjustment recommendation. The "tighten spiral" is self-perpetuating until the root cause is addressed.
CPAP fitting intake should ask whether the patient shares a bed. Exhaust direction is a variable that most selection processes never surface — but it directly affects therapy adherence for a meaningful patient segment.
The most consistent finding across all six: the problems driving mask failure are specific, identifiable, and addressable. What most patients lack is the framework to diagnose them — not the ability to solve them.
What the data says, in plain language
- 72% of patients try more than one mask — this is normal, not a failure
- The Ghost Leak affects 49% of all users and is almost never caused by the mask model
- Changing masks resolves a Ghost Leak only 37% of the time — the cause usually travels with the patient
- Side sleepers are twice as likely to still be searching, and have a tube problem in addition to a seal problem
- Finding the perfect mask doesn't end overnight disruption — tube management and cushion maintenance are separate, addressable problems
- Bridge pain and active leaking coexisting is almost always a cushion-size issue, not a strap-tension issue
- Exhaust comfort for bed partners belongs in the fitting conversation for patients who share a bed
Try MaskFit AR!
The fitting tool built in response to this data
MaskFit AI is a pre-fitting screening tool available exclusively through The CPAP Shop. It was built directly in response to the patterns in this data: the disproportionate failure rate among side sleepers, the persistence of Ghost Leaks across mask changes, the cushion-size errors driving most bridge pain cases.
The result: patients arrive at fitting appointments with a shortlist of masks that address their specific pattern, rather than a generic "best masks" ranking. Average masks fitted to average faces produce the average outcomes our data documents. MaskFit is built for everyone else.









