If you market a sleep clinic in Australia, 2 regulators watch your ads. And the diagnosis you sell is usually made by someone else. Medicare pays for a home sleep study only when a qualified adult sleep medicine practitioner reads it, looking at the raw data in person. The National Law covers how you advertise the service. The Therapeutic Goods Advertising Code covers the CPAP machine. I read 31 Australian sleep clinic websites. Of those, 27 offer a sleep study, and 8 of the 27 never say who reads it.
Who makes a sleep apnoea diagnosis in Australia?
A sleep physician makes the diagnosis. The clinic that wires your patient up usually does not.
The clearest statement of this is in the Medicare Benefits Schedule. Item 12250 is the home sleep study. Medicare calls it an unattended study, and its purpose is "to confirm diagnosis of obstructive sleep apnoea". The schedule fee from 1 July 2026 is $401.25. The attended laboratory study, item 12203, is $703.80.
Item 12250 comes with conditions, and 2 of them matter for your website.
- Who decides the study is needed. It is a qualified adult sleep medicine practitioner or a consultant respiratory physician. They can decide from a GP referral with screening scores, or after seeing the patient.
- Who reads the study. The item requires that "interpretation and preparation of a permanent report is provided by a qualified adult sleep medicine practitioner with personal direct review of raw data".
The screening scores are set out too. The patient needs a STOP-Bang score of 3 or more, an OSA50 score of 5 or more, or a high risk Berlin score. They also need an Epworth Sleepiness Scale score of 8 or more. The notes say your sleep technician or other staff may give the questionnaires. The physician then uses the results.
So the work is split. A technician fits the sensors. A sleep scientist scores the recording. A physician reads it and makes the diagnosis. Many sleep clinics do the first step, some do the second, and only some employ the third.
That split is the whole marketing problem. Your patient wants a diagnosis, yet many clinics sell a recording.
Which advertising rules cover a sleep clinic website?
Two sets of rules apply, and most sleep clinics fall under both.
The National Law covers the service. Section 133 is the part of the law that sets the health ad rules. It applies when you advertise a registered health practitioner's service. Ahpra's advertising guidelines list 5 things your ads must not do. They must not be false, misleading or deceptive. They must not offer a gift or discount without the terms. They must not use testimonials. They must not make people expect more from treatment than is fair. And they must not push people to use health services they do not need.
Ahpra's page also gives the penalties. The National Law changed in 2022. The maximum is now $60,000 per offence for a person and $120,000 for a company. My guide to the Ahpra advertising guidelines covers all five bans.
The Therapeutic Goods Advertising Code covers the device. A CPAP machine, a mask and a mandibular splint are medical devices. The Code applies to anyone who advertises them to the public. You do not need to be a registered practitioner for it to apply to you. A CPAP shop with no clinician on staff still has to follow the Code. My guide to the TGA Advertising Code has the full list of rules.
Hearing clinics have the same shape, with 2 regulators. A regulated service is wrapped around a regulated device. I measured that in the audiology marketing post. Sleep adds a third layer, which is Medicare deciding who may make the diagnosis.
What did I find on 31 sleep clinic websites?
I searched for "sleep clinic" plus a city, across Sydney, Melbourne, Brisbane, Perth, Adelaide, Canberra, the Gold Coast, Hobart and Newcastle. I used no claim words in any search. I left out public hospitals, state health services, research institutes, directories and overseas results.
That gave 39 private clinic sites. Of those, 7 would not load and 1 showed no readable text. I read the other 31. That came to 161 pages and 108,418 words, counting repeated menus and footers once. Of the 31, 27 offer a sleep study themselves. The other 4 treat or assess without one.
Source. My own reading of 31 Australian sleep clinic websites on 21 September 2026. A keyword scan found the candidates and I confirmed each one by hand. The sentence behind every count is on file. No clinic is named.
Here is the good news first. Of the 27 clinics that sell a study, 19 do say who reads it. The clearest of them are blunt about it. One tells patients to see the specialist who reported their study, and adds: "Don't get your results from the technician who wired you up". Another explains its cheaper screening test honestly. Those results, it says, "rely on software for making a correct prediction", with no scientist or physician reading them.
That kind of sentence is rare. On your page, it is worth more than any slogan.
Why does it matter if the page never says who reads the study?
If you leave a gap, your reader fills it with the wrong answer.
Say your page reads "we diagnose sleep apnoea" next to a photo of a friendly technician. Your patient will think the technician makes the diagnosis. If a physician in another city reads the file, your page has left out the most important fact about the service. Section 133 bans ads that mislead or are likely to mislead. Consumer law says the same for any business. Leaving out who makes the diagnosis is the kind of gap both rules are aimed at.
Of the 27 sites that offer a study, 8 never close that gap on the pages I read. Some of those 8 are physician practices, where a specialist plainly does the reading. They still do not say it. A patient comparing 5 tabs cannot tell a clinic led by a physician from a device shop. Your page has to tell them.
Two lines from the sample show how the gap opens.
One clinic offers "home sleep studies with or without a sleep physician consultation". That fairly describes 2 options. But it also tells your reader that the physician is optional. If the Medicare pathway is what you sell, say who reports every study, consult or no consult.
Another page tells readers how simple the diagnosis is. If a partner has noticed gasping or choking, it says, "then sleep apnoea is present". Now a web page is making a diagnosis, for a reader nobody has assessed. A specialist wrote it, and the clinical point may be sound. But as an ad, it tells a stranger they have a condition.
What can you say about a diagnosis you do not make?
You can say everything that is true about your part of the work. The fix is to name each step and who does it.
| The pattern | Why it is a problem | Use this instead |
|---|---|---|
| We diagnose sleep apnoea. | The reader cannot tell who "we" is. If a physician elsewhere reports the study, the line misleads. | We fit and run your home sleep study. A sleep physician reads the recording and makes the diagnosis. |
| Get diagnosed at home tonight. | It promises a result and a speed. The recording happens tonight. The report does not. | You do the recording at home in one night. The physician's report usually takes [your real turnaround]. |
| Find out if you have sleep apnoea with our free quiz. | A screening score is a risk estimate. Medicare itself treats it as a step before the study. | This quiz estimates your risk. It cannot diagnose sleep apnoea. Only a sleep study read by a physician can. |
| Bulk-billed sleep studies. | Item 12250 has conditions. A reader who does not meet them will be charged. | Bulk billed if you meet Medicare's criteria, which include a GP referral and screening scores. We check this before you book. |
| Untreated sleep apnoea leads to heart attack and stroke. | Fear sells studies. Section 133 bans encouraging unnecessary use of health services. | If you snore and feel tired most days, ask your GP whether a sleep study makes sense for you. |
Fill the square brackets with your own number before you publish. If you state a turnaround you cannot keep, that is a misleading claim in itself. In the sample, the stated turnarounds ran from "7 to 14 days" to "two to four weeks".
When does a CPAP page need a mandatory statement?
Your CPAP page needs one when it does more than list the product.
Section 20 of the Code says what an ad for a medical device must contain. It needs the trade name, an accurate description and at least one accepted purpose. It also needs one of 2 statements, shown clearly. They are ALWAYS FOLLOW THE DIRECTIONS FOR USE, or ALWAYS READ THE LABEL AND FOLLOW THE DIRECTIONS FOR USE. Say your page lets people buy the device without seeing it first, as any online shop does. Then you must also list or link the health warnings.
One exception matters here. Section 14 covers an ad that shows only the product name, a picture, the price and where to buy it. If it makes no health claim, the statement rules do not apply. So a bare product listing is exempt. But once your listing says the machine treats sleep apnoea, it is no longer exempt.
Of the 31 sites, 6 name a device brand they supply. None of the 31 carry either set statement on any page I read. I did not open every product page, so I cannot say which listings needed one. I can say the statement was nowhere on 161 pages, including pages that describe what CPAP treats.
There is a second Code issue, and it is a bigger one. Section 28 explains when a condition counts as serious. One test is that "there is a diagnostic (including screening), preventative, monitoring, susceptibility or pre-disposition test available for the form (including a self-administered test), which requires medical interpretation or follow-up". A sleep study is a test that needs a doctor to read it. Medicare says so in item 12250. On a plain reading, sleep apnoea meets that test. If your device ad mentions a serious condition, the TGA calls that a restricted representation. You need TGA approval or permission before you use one.
That is my reading of the text, and it is general information. Before you build a CPAP sales page around the words "treats sleep apnoea", check the current permissions with the TGA or a lawyer who knows these rules.
Can a sleep clinic publish patient reviews?
You can, but it depends on what the review talks about and which regulator is looking.
Under the National Law, a testimonial is a positive comment about the clinical side of your service. Ahpra's guidelines say a review is clinical when it mentions a symptom, a diagnosis, a treatment or a result. A review about your friendly staff and easy parking is fine. My post on patient testimonials goes through the line in detail.
Of the 31 sites, 7 publish patient reviews on their own pages. On 3 of them, a review describes a diagnosis, a device or a result. One says it "was confronting to be diagnosed with sleep apnea". In the same set of reviews, a patient names the CPAP model they bought. One says "I am 100% better than I was".
Those are warm, real-sounding words, and that is the trouble. They are exactly what section 133 means by a testimonial.
The device side has its own rule. Section 24 of the Code says a testimonial about a health product must describe a typical health benefit. You must check what it says and who wrote it. And an endorsement must not come from "a current or former health practitioner, health professional or medical researcher". So say a sleep physician praises a named splint on the page that sells it. That breaks the Code, even where the National Law says nothing.
Is an online sleep apnoea quiz advertising?
Yes, it is, if it sits on your site and ends in a booking button.
Of the 31 sites, 6 offer a quiz or risk test. Only 2 of those 6 say the result is not a diagnosis or does not replace a sleep study, on the pages I read. One does it well: "This is a basic screening tool for Obstructive Sleep Apnea and not a medical diagnosis."
The quizzes themselves are sound. STOP-Bang, OSA50 and the Epworth scale are the same tools Medicare names. The risk lies in how you wrap them. A headline that asks "Do you have sleep apnoea?" and a result page that says "high risk, book now" turns a screening tool into a sales funnel. That is where the ban on "indiscriminate or unnecessary use" comes in. It means pushing people towards care they may not need.
You can fix most quizzes with 3 lines. Say what the tool is and what it cannot do. Then say who decides what happens next.
How did I measure this, and what can it not tell you?
I ran 9 neutral searches on 21 September 2026 and took the private clinic sites they returned. From each site I read the page the search led to, plus up to six more: the home page and linked pages about sleep studies, sleep apnoea, CPAP, snoring, treatment, services, FAQs and testimonials. The same rule ran on every site.
A keyword scan found candidate sentences. I then read them and classified each site by hand. A site counts as saying who reads the study only if a sentence ties its own study to a physician, specialist or doctor who interprets, reports or diagnoses from it. General education lines do not count.
- 39 sites found, 7 would not load, 1 had no readable text. The sample is 31 sites and it is small.
- I read up to seven pages a site. A clinic may name its reporting physician, or carry a mandatory statement, on a page I did not open.
- A scanner cannot read text inside images, PDFs, videos or booking widgets.
- I did not check who reports each clinic's studies in real life. I checked what the website says.
- The section 28 point is my reading of the Code's words. I could not load the TGA's own guidance page on restricted representations, so I have not relied on it.
- This is general information about advertising rules. It is not legal advice about your practice.
What should you check on your own site this week?
Open your sleep study page as if you were a stranger, and look for 5 things.
- The reader. Find the sentence that says who reads the recording. If it is missing, write it. Name the role, and the person if you can.
- The verb. Search the site for "diagnose". Check that the subject of each sentence is the person who really does it.
- The quiz. Check the result screen. It should say the score is a risk estimate and that a physician decides.
- The reviews. Read each one for a symptom, a diagnosis, a device or a result. Take those down.
- The device pages. If a page says what a machine or splint treats, it needs the mandatory statement, and it may need more than that.
If you want the whole site read this way, with every flagged line quoted and the replacement written, that is what the All Clear Audit does. If you would rather start with the pages patients read first, website copy is the place. It is written by me, Amisha Sharma, at Commas & Chaos.
Prefer the short version? The findings are in an eight-slide web story.
Related reading:
