PSG Codes, CPAP Setup & HST Billing Rules
Sleep medicine billing isn't particularly difficult, but it's got a handful of rules that catch practices off guard, especially around split-night studies, home sleep test timing, and CPAP supply billing. Miss one and you're looking at a denial that takes three times as long to appeal as it would've taken to get right the first time. Get this wrong at volume and the cost shows up fast, through denials, undercoding, or missed follow-up.
Here's a practical walkthrough of what applies in 2026, from the diagnostic study through PAP therapy management.
If you're working through these issues, our specialty billing services can help you address them systematically.
| CPT | Description | Type |
|---|---|---|
| 95810 | Polysomnography, attended, age 6+, sleep staging + 4 or more additional parameters | In-lab PSG |
| 95811 | PSG with CPAP titration, attended | In-lab titration |
| 95782 | PSG, attended, younger than 6 years | Pediatric |
| 95783 | PSG with CPAP, younger than 6 years | Pediatric titration |
| 95800 | Home sleep apnea test, unattended, Type III | HST |
| 95801 | HST, minimum 3 channels | HST |
| 95806 | HST, unattended, Type II or IV | HST |
The attended vs. unattended distinction matters. A technologist must be present for attended PSG codes (95810, 95811). Home sleep tests are unattended by definition, the patient takes the device home and records independently.
A split-night study is when the first half of the night is used for diagnostic recording and the second half for CPAP titration, all in one encounter. You bill 95811 for the entire night. Not 95810 plus 95811. One code.
CMS requires at least 2 hours of diagnostic recording and an AHI of 40 or greater (or 20+ with complicating factors) before switching to titration mid-night. Document the exact times and the AHI that triggered the protocol change. Payers audit this.
Most commercial payers and Medicare prefer HST over in-lab PSG for straightforward adult patients with suspected obstructive sleep apnea. It's cheaper for them, and for uncomplicated cases, the diagnostic accuracy is adequate.
The billing nuance: who owns the HST device matters. If your practice owns the device and loans it to the patient, you can bill for the technical component. If the patient rents a device from a DME supplier, the DME company bills for the device and your practice bills only for the interpretation (modifier 26).
HST is not appropriate, and payers will deny if it's used, when the patient has significant comorbidities: congestive heart failure, COPD, neuromuscular disease, or suspected non-OSA sleep disorders like central sleep apnea, periodic limb movement disorder, or parasomnias. Those patients need full attended PSG. Document the clinical reason for ordering in-lab rather than home testing or you'll spend time defending an unnecessary in-lab denial.
| CPT | Description |
|---|---|
| 94660 | CPAP initiation and management |
| 95800–95806 | Download and review of PAP device data (use HST codes or PAP adherence codes depending on context) |
| 99091 | Collection and interpretation of physiologic data, 30+ min/month (for remote PAP monitoring) |
94660 is billed by the physician or qualified provider for the face-to-face encounter when setting up or managing CPAP therapy. It's a professional service, not a facility or equipment code. The DME supplier bills separately for the physical CPAP device under HCPCS E0601.
CPAP supplies (masks, tubing, filters, chinstraps, headgear) bill under HCPCS codes through a DME supplier, not through the physician practice. If your practice is also functioning as a DME supplier, you need a separate DME supplier number and billing system. Practices that try to bill CPAP supplies through their professional billing without DME accreditation run into immediate denial and potential compliance issues.
Virtually every commercial payer requires prior auth for in-lab PSG. Many also require it for HST, though the bar is lower. The auth criteria almost always include a clinical questionnaire (Epworth Sleepiness Scale, STOP-BANG) and documented symptoms.
The denial pattern we see most often: a sleep physician orders a PSG after a clinical visit, the auth is submitted with the CPT code but without the supporting documentation, and the payer denies for lack of medical necessity. Fix this by attaching the questionnaire scores and the clinical note with the auth submission; don't make the payer ask for it.
If your practice reads studies from an outside sleep lab, you're billing only the professional component (26 modifier). If you own your own lab and perform both the technical and professional components, you bill the global code. Don't bill global if you don't own the facility; that's a compliance problem, not just a billing error.
See how Rcmaxis handles the full scope of Sleep Medicine billing on our Sleep Medicine billing services page.
Sleep claims are reviewed for the type of study and whether it was appropriate for the patient. Payers prefer home sleep testing for straightforward adult cases, so an in-lab study for a patient without a documented reason to go to the lab is a common denial. Split-night studies are checked to confirm the whole night was billed as one study and not as two. A 3-year lookback with even a modest error rate can turn a small per-claim gap into a sizable repayment demand, which is why a quarterly self-check is worth the time.
CPAP follow-up matters because PAP therapy coverage usually depends on documented adherence. Contractors can look back three years, and equipment and supply claims that rest on missing adherence data are a frequent repayment finding. Keep the clinical questionnaire, the study report, and the adherence download together in the chart so the whole story is in one place.
A sleep study touches several steps: the order and the authorization, the scheduling, the test, the scoring, the interpretation, and the claim. Follow each study through a simple tracker so that nothing sits between steps. Check that the authorization matches the study type performed, since a switch from an in-lab study to a home test, or the reverse, may need a new approval.
At the end, confirm the report is signed and the code matches the study. Studies with missing or unsigned reports are among the easiest claims to lose.
Sleep studies are high-value claims. At 2026 national rates, an attended in-lab polysomnogram (95810) pays $673.70, a split-night study with CPAP titration (95811) pays $707.77, and a home sleep test (95806) pays $103.21. A lab that runs 40 attended studies a month bills $26,948.
If 8% of those claims deny for a missing authorization, that's about 3 studies and $2,156 a month. Most payers won't pay retroactively for a study done without approval, so the check has to happen before the patient arrives. Confirm the authorization number, the approved code and the date range at scheduling, and again the day before.
As one code. The article says you bill 95811 for the entire night when the first half is diagnostic and the second half is CPAP titration, not 95810 plus 95811.
For straightforward adult patients with suspected obstructive sleep apnea, it costs less and the diagnostic accuracy is adequate, according to the article. An in-lab study usually needs a documented reason.
When your practice reads studies from an outside sleep lab, you bill the professional component with modifier 26. If you own the lab and perform both parts, you bill the global code.
The billing gaps most practices don't catch until they show up as denials. Get the checklist: free, no spam.
✓ Got it! Your info is saved: view the checklist now →
We don't share your info. Unsubscribe any time.
Revenue check: A quarterly sample of 20 to 30 claims here, checked against the supporting documentation, is cheap insurance against a $15,000 to $35,000 repayment surprise down the line.
We specialize in sleep medicine billing, from PSG auth workflows to PAP adherence documentation. Let's find what's costing you.
Get Your Free Revenue Assessment