Beginning of the End for Polysomnography?
The AASM released a “Clinical Practice Guideline” for diagnostic testing for OSA.1 Not surprisingly, the guideline is heavy on home sleep apnea testing (HSAT). This replaces a practice parameter and a clinical guideline from a decade ago. I think it also supersedes the SCOPERS paper technology review. It provides good practice statements for evaluating patients and using a PSG as the standard diagnostic test. There are recommendations for initial evaluation of risk; HSAT device type and sensors; a definition of a technically adequate test; and recommendations on what to do when the HSAT is negative.
A Blueprint for an HSAT Program?
The most revealing section is the “Remarks” section accompanying recommendations 2 and 3. The remarks are “based on specifications used by studies that support these recommendation statements.” (p. 490) I view this as the recommended blueprint for an HSAT program, and yet the task force labels this as “Remarks,” puts the whole thing in italics, and buries the whole thing 11 pages into the paper.
Previous publications gave wide latitude to sleep programs, including which questions should be used in screening patients, what devices are acceptable, which are appropriate sensors and what are the standard protocols. The vagueness left centers wondering which to choose. Should I use the Berlin or the OSA-50? Which is better: thermal sensors or RIP belt sums? Do I dare to eat a peach? (T.S. Eliot reference for poetry nerds.)
How to Run an HSAT Program
Here is how to run an HSAT program based on the remarks:
- Identify “uncomplicated” patients by the absence of symptoms of non-obstructive breathing, the absence of non-respiratory sleep disorders, and the absence of personal factors that make a successful HSAT study unlikely
- Identify patients at risk for moderate to severe OSA based on EDS plus two of these three symptoms: loud snoring, witnessed apnea or diagnosed hypertension
- Be an accredited center supervised by a board-certified sleep specialist
- Run patients for a single night
- Use devices that record, at a minimum, nasal pressure, two RIP belts, and a pulse oximeter (alternatively use a PAT device with oximetry and actigraphy)
- Run studies for at least four hours
I’m impressed that the authors were willing to take such a bold step. Now we have a good idea of what to do. The authors are careful not to say that this is the only acceptable pathway.
However, if you are starting a program and looking for an evidence-based approach, the subtext is that this is the way to go. If you cut and paste the above into your policy and procedures manual, it would be hard for someone to say that you are doing it wrong.
Nasal pressure sensors are favored over thermal sensors or other flow surrogates. This makes sense as nasal pressure is recommended for detection of hypopnea. RIP belts are more reliable than piezo-belts and don’t reverse polarity in mid-recording. A single night lasting at least four hours sounds like a fair sample of data.
Beginning of the End for Polysomnography
So why do I say it is the beginning of the end for polysomnography? The clinical practice guideline still says PSG is the gold standard for diagnosis of OSA. I have three reasons:
- The clinical practice guideline defines how HSAT should be used and I am sure the field will quickly coalesce on this standard methodology.
- Now that there is a standard methodology it will be tested on the complicated patients that are currently excluded from HSAT use. They are only excluded because there is no data to show accurate estimation of AHI. It shouldn’t take long for that to accumulate, especially since the recommended sensors are the same as for the respiratory portion of the PSG. My guess is that HSAT will be adequate for diagnosis of OSA in the clear majority of patients, including those with non-obstructive apnea and comorbidities.
- It is estimated that “the ratio of people to sleep specialists in the United States is more than 43,000:1”2 It is unlikely that the needs of patients with sleep disorders will be met by sleep specialists soon. If primary care physicians are diagnosing OSA, it is far more likely that they will be using HSAT than PSG.
But, you may be thinking, there’s still narcolepsy. You can’t diagnose narcolepsy without a PSG and MSLT. For now. Let’s ask Dr. Mignot if he has a narcolepsy blood test ready to go yet.
And this is just the start. Wearable technology, under-the-mattress sensors and touch-free devices will soon be as accurate as HSAT and supported by the same weight of evidence. These factors are why the AAST has recommended that technologists shift their emphasis from diagnosis to treatment.
The rise of HSAT will result in sizeable increases in the number of patients diagnosed with OSA. Physicians won’t be able to see and monitor all of them. As the number of PSGs decreases, technologists will be perfectly positioned to rotate their sleep schedules and spend the day putting patients on the road to healthier sleep. The future’s so bright, you’ll have to wear shades.
Interested in learning a perspective on where sleep technology is today and where we see its headed in the future? Just click on the image below and get a FREE learning module on this very subject!
1Kapur VK, Auckley DH, Chowdhuri S, Kuhlmann DC, Mehra R, Ramar K, Harrod CG. Clinical practice guideline for diagnostic testing for adult obstructive sleep apnea: an American Academy of Sleep Medicine clinical practice guideline. J Clin Sleep Med. 2017;13(3):479–504.
2Watson NF, Rosen IM, Chervin RD, Board of Directors of the American Academy of Sleep Medicine. The past is prologue: The future of sleep medicine. J Clin Sleep Med. 2017;13(1):127–135.