Sleep study billing services sit on a split that looks simple and causes constant trouble. A study is either attended by a technologist or it is not, and that single fact determines the code family, the place of service, the payment, and in the case of Medicare, the entire code set.
Getting the split wrong is not a minor coding matter. Billing an unattended home study as attended polysomnography is upcoding, and it carries exposure well beyond a denied claim. Billing an attended study under home testing codes leaves a large amount of earned revenue unclaimed. Sleep programs lose money in both directions, often at the same time.
Attended & Unattended Are Different Services
If a qualified technologist monitored the patient throughout the study, the service is polysomnography. If nobody was present, it is an unattended sleep study regardless of how many channels the device recorded or how thorough the report reads.
Attended studies are expected to be performed in a facility based sleep laboratory rather than in the home or a mobile facility, and the documentation has to show that. Unattended studies may be performed in the home or, for certain code families, in a facility.
Channel count does not convert an unattended study into polysomnography. That is worth stating plainly, because the assumption that a well-instrumented home device qualifies is a recurring source of trouble.
In-Lab Polysomnography
For patients six years and older, the adult family covers diagnostic polysomnography with sleep staging and four or more additional parameters, polysomnography with positive airway pressure titration, and a lower-parameter option covering one to three additional parameters. Patients under six have their own diagnostic and titration codes.
Split-night studies are the ones that get miscoded. When a diagnostic study converts to titration in the same night, the titration code is billed alone, since it includes the diagnostic portion. Reporting both produces a denial and looks like duplicate billing.
Two bundling rules save practices from denials they create themselves. The polysomnography codes include sleep staging, so separate billing for EEG, EOG, or EMG alongside them is not expected. The codes also include their stated parameter counts, so additional parameters are not separately reportable.
Studies with less than six hours of recording, or otherwise reduced in scope, are reported with the reduced services modifier rather than billed as complete.
Home Testing Splits by Payer
This is the part that surprises programs expanding into home testing, and it has been true since 2008.
Medicare does not accept the CPT codes for home sleep apnea testing. Home studies for Medicare patients are billed under the G-code family, and submitting the CPT codes instead produces automatic denials. Commercial payers vary. Some accept the G codes, others require the CPT codes, and there is no way to know which without checking each contract.
Running one code set across all payers guarantees denials on part of the book. The workable answer is a payer-by-payer matrix showing which code set each contract accepts, because the answer does not follow any pattern that can be reasoned out from the outside.
Device Type Drives Code Selection
Within the CPT family, the correct code depends on which channels the device actually recorded. Devices using peripheral arterial tone rather than airflow do not fit the code that requires respiratory airflow measurement, and coding by device brand habit rather than by recorded channels produces errors that survive for years.
Consecutive Nights Are One Test
Multiple consecutive nights of home testing are treated as a single test. Additional payment is not made for the extra nights. Programs that bill each night separately are generating denials and, if the pattern persists, something worse.
When more than two nights of testing are performed, the medical necessity for the additional nights has to be documented in the record.
The Codes Change on January 1, 2027
Sleep programs have a limited runway on this one. The AMA CPT Editorial Panel approved a new code set for home sleep testing at its February 2025 meeting and approved deletion of the three current unattended study codes, with the change effective January 1, 2027.
The in-lab polysomnography codes are not affected. Programs with significant home testing volume should be planning the transition now, including system updates, payer communication, and confirmation of how each contract will handle the new set. Practices that wait until the codes disappear will spend the first quarter of 2027 working rejections.
Coverage Conditions
Medicare coverage for sleep testing in obstructive sleep apnea runs through national coverage determinations, with contractor articles adding the coding detail.
Home testing is expected to be performed alongside a full clinical sleep evaluation, in patients with a high pretest probability of moderate to severe obstructive sleep apnea. A home test ordered without that evaluation on record is exposed even when the result is positive and the treatment that follows is appropriate.
Diagnosis coding has to support the study. The obstructive sleep apnea diagnosis is the primary code on most of these claims, and studies ordered for symptoms alone need the symptom coding to line up with the covered indications in the applicable article.
Where the Money Leaks
Denials in sleep medicine cluster in predictable places. Home studies billed to Medicare under CPT codes rather than G codes. Split-night studies billed as two services. Separate billing for components already included in the polysomnography code. Multiple nights of home testing billed as multiple tests. Studies under six hours billed without the reduced services modifier. Home tests performed without a documented clinical evaluation preceding them.
There is also a quieter loss. Sleep programs frequently bill only the technical component when they own the equipment and a physician in the group performs the interpretation, or the reverse, without confirming which components they are entitled to report. That gap does not generate a denial. It simply pays less than the work supports, and reviews of sleep program billing, including the specialty audits run by firms such as AAA Medical Billing, tend to surface it late because nothing in the remittance points to it.
What Holds Up
The reliable safeguard is a check at the point of coding that answers three things: who attended the study, what the device actually recorded, and which code set this particular payer accepts. Those three answers determine the claim.
Add the 2027 transition to the planning calendar now rather than in December. The code change is settled, the effective date is fixed, and the programs that handle it early will be billing normally while others are working through rejections.