Do I Really Need an Overnight Sleep Study (Polysomnogram)?: A ReLACSing Blog #29
The infamous 20th century criminal Willie Sutton, when asked by a reporter why he robbed banks, answered, “Because that is where the money is.” I could easily end this blog quickly by saying that this is the very same reason you really do need an overnight sleep study (polysomnogram or PSG for short). The PSG is exactly where the money is for the typical insurance-based sleep center, so you better believe this is a very large reason why so many PSGs are ordered. As you can imagine, with money involved, the ordering of the PSG does not always overlap with the need for one. This blog explores the actual medical reasons why you may need an overnight sleep study.
First, what is a polysomnogram? Getting back to its Greek and Latin roots, polysomnography means “multiple sleep writings.” It is a diagnostic procedure that measures various aspects of sleep and bodily function that originally consisted of multiple recording channels continuously printed out on paper, the “graph” part of the word. Now done via computers with human interpretation (and very soon AI), it is set up by a sleep technologist who sticks a bunch of electrodes and other sensors all over the body and then somehow the individual will sleep so that measurements can be obtained. PSGs measure brain waves, eye movements, muscle activity in the chin, legs, and sometimes arms, heart rhythm, oxygen levels, and potentially more complicated things like carbon dioxide (CO2) monitoring. The PSG tracks breathing using elastic belts around the chest and belly, as well as various tubes in and around the nose and mouth for airflow. A person having a PSG will check into the sleep laboratory, get wired up, sleep for one night, and then go home.
Because insurance favors things you do to someone (procedures) over the quality of what you do with someone (e.g. a doctor spending more time talking to a patient), this procedure in sleep medicine is the go-to test for the field. It involves substantial overhead, lots of equipment, and human labor to set up, monitor, and interpret. Thus, insurance can be charged a few thousand dollars with a lot of margin to profit the center. A home sleep apnea test (HSAT) is a lesser, simpler test to diagnose obstructive sleep apnea. It involves just a few sensors and is done at a person's home. It is reimbursed significantly less than a PSG, with reduced profit margin. Think of Ford Motor Company and their decision to ax the Escort (and more recently, the Focus) in favor of the mighty F-150. In fact, the Escort used to outsell the F-150 in volume but not in profit margin. The HSAT is a Ford Escort and the PSG is an F-150. However, Ford convincing people to buy their cars is really not the way in which sleep centers should be selling patients on getting sleep studies, but it sometimes feels that way.
With the conflict of interest here duly noted, just when would someone really need an overnight sleep study? Answer: as often as you can get one approved by insurance! OK, let’s restate that. Just when would someone really need an overnight sleep study for medical reasons? Let’s investigate.
The primary reason for a sleep study is to diagnose obstructive sleep apnea (OSA). OSA is probably the most common sleep condition, increasing rapidly in prevalence in the US due to the obesity crisis. OSA is a condition in which the upper part of the throat becomes weak or too narrow and begins to collapse during sleep when the muscles relax or go out completely in REM sleep. To the disturbed spouse, this may result in a snoring sound, but the bigger problem is that partial or complete blockage of airflow can lead to frequent awakenings and drops in oxygen levels that can create a cascade of effects on daytime function, alertness, cardiovascular health, and numerous other health factors. Home sleep apnea testing can be used to diagnose the vast majority of cases of OSA. Some traditional devices measure airflow, oxygen, snoring, and usually body position. Other HSAT technologies detect sleep apnea by indirect measurements including arterial tone and heart rate. The technology continues to improve, and hopefully one day an HSAT will be able to measure sleep itself with nearly as much detail as a PSG. Right now, however, home testing really cannot measure sleep, just like wearable devices are still struggling in that arena.
An HSAT should not be used to diagnose OSA in some circumstances as it does not provide the level of information to make diagnosis and treatment decisions for more complex conditions that may overlap with OSA. Here are many of the medical reasons for getting the more detailed PSG over the HSAT:
Heart failure: Those with significant heart failure or other disruption to heart function may need a PSG due to a high risk of central sleep apnea (CSA) as described in this video. This form of sleep apnea is different from OSA, and it involves pauses in breathing or an irregular breathing pattern. OSA is where you try to breathe and your throat obstructs, but CSA is where you may breathe too fast, too slow, or stop breathing entirely for many seconds. This is quite complex and involves the consistency of blood flow to the breathing centers of the brain. Distinguishing between OSA and CSA is best done with the more detailed PSG as these two things can be hard to distinguish with the limited amount of information provided by an HSAT.
COPD, significant lung disease, and those on supplemental oxygen: Problems with lung performance and breathing when one is not sleeping can create additional problems in combination with OSA when one is sleeping. The PSG provides more detail for breathing episodes and gives the option for CO2 monitoring, which can be a problem in COPD or other lung diseases. Sometimes persistently low oxygen levels on an HSAT can be indistinguishable from sensor error (called artifact) from the unit, which is unmonitored at a person’s home.
Severe obesity (BMI > 50 kg/m2): Though a person with significantly increased body mass may likely be diagnosed with sleep apnea through an HSAT, it may be of use to perform a PSG in order to monitor CO2. Those with a very high BMI may be at risk for hypoventilation (under-breathing) with decreased oxygen and increased CO2 levels without actually having OSA or in addition to the OSA. However, one could make an argument that you can still use an HSAT to diagnose a breathing problem and then have it treated appropriately in an overnight study to test CPAP (addressed below).
High doses of opioids: Those who are high doses of full agonist opioids (e.g. oxycodone, methadone, fentanyl but not buprenorphine) could develop CSA and/or hypoventilation from the effects of opioids on the breathing system controlled by the brain. A dramatic extension of this concept is an opioid poisoning from fentanyl or heroin on the street that can impair breathing so much that the individual may die, which is why there are so many opioid-related deaths from poisoning in the US currently. Like in heart failure and severe obesity, the HSAT may not be as good as the PSG at providing all details of breathing ability.
Spinal cord injury: similar to above, individuals with significant injury to the spinal cord in the upper back or neck can also be at risk for CSA and/or hypoventilation and should have in-lab sleep testing.
Negative, borderline, or inconclusive HSAT: for those in which the HSAT is insufficient, a PSG is needed. Home sleep apnea testing is not sensitive for diagnosing milder or borderline forms of OSA. It also cannot measure sleep, so if a person has significant insomnia during an HSAT, they may look like they don’t have sleep apnea because they were awake and breathing normally most of the study. With the HSAT, you can’t tell. If there is suspicion for OSA and the HSAT cannot confirm it, most people need a PSG to be sure.
Diagnosis of narcolepsy type 1 (NT1): as I mentioned in a two-part blog (read parts 1 & 2 here) on this condition, polysomnography can be helpful in the diagnosis of NT1, but it is not essential. The absence of hypocretin in the cerebrospinal fluid is the definitive diagnosis. However, most sleep clinics are not measuring hypocretin levels and end up doing a lot of PSG as well as its companion, the multiple sleep latency test (MSLT), for bigger margins and with often good coverage from third-party payers. Read those previous blogs for more information as well as this one on idiopathic hypersomnia and this one on narcolepsy type 2.
Aside from sleep apnea and sleep-disordered breathing, there are a few other sleep conditions that warrant polysomnography:
REM sleep behavior disorder (RBD): Those with dream-enactment behavior need a PSG to be monitored for how their rapid eye movement (REM) sleep looks. The diagnosis of RBD can be confirmed in those who have abnormal muscle activity or even are observed in the sleep study video to act out dreams during REM sleep. HSATs do not measure sleep at all, much less muscle tone during specific stages of sleep and is unhelpful for this condition.
Sleepwalking and other behaviors that occur while sleeping, also called parasomnias: RBD is an example of a parasomnia, or something that happens alongside sleep (coming from the Greek and Latin roots). Other parasomnias are quite common like sleep-talking and sleepwalking, but some of these behaviors require identification, diagnosis, and treatment. Polysomnography allows clinicians to capture the events, determine the sleep states and stages in which they are occurring, and see if they are triggered by certain things like sleep apnea.
Seizures: These can occur while sleeping and must be distinguished from parasomnias. Sometimes more extensive brain wave monitoring can be performed, identical to the numerous EEG electrodes placed on the scalp that are used in seizure clinics or in the hospital.
Those who have been diagnosed with OSA, even after a home sleep apnea test, will often be told that they must have an in-center sleep study to test CPAP. The true wording should be sometimes can instead of must. (Often PAP titrations are covered by insurers when PSGs are not, so if you didn’t ring the register the first time, here’s a second chance, ka-ching!)
Continuous positive airway pressure (CPAP) is by far the most common and effective treatment for OSA. For those who have not heard of this device, it is basically a reverse vacuum connected by a tube to a mask on one’s nose that puts pressure in the upper airway to prevent the collapse, thus eliminating the breathing episodes (and snoring!). CPAP came out in the early 1980s, and there have been many advancements including the advent of automatic CPAP machines (APAP, autoPAP, or auto CPAP), which have eliminated the need to test the settings of the CPAP in most patients. It used to be that you had one pressure setting and it had to be determined by the PSG, but now that is only true in specific cases but not for the majority. These overnight sleep studies are called PAP titration studies. Their margins are even slightly higher than regular PSGs because you are doing even more than just monitoring sleep. When are PAP titrations actually indicated?
Most of the same reasons as above: Depending on the specific medical conditions of an individual, the reasons to get a PSG over an HSAT apply to PAP titration. However, some PSGs can show enough to eliminate the need for a PAP titration and instead move to an autoPAP machine. For example, a person with heart failure, opioids, COPD, or significant body mass can have a PSG showing only straightforward OSA without any CSA, low oxygen, high CO2, etc. and do quite well on autoPAP with the risks investigated and ruled out from the first study.
Significant central apnea: As addressed in this video, not all CSA is pathological. However, significant central sleep apnea seen on an HSAT or PSG necessitates an in-lab titration study to test PAP therapy and potentially multiple device settings or the addition of supplemental oxygen. AutoPAP rarely can impact CSA and really only treats OSA.
Hypoventilation or persistently low oxygen (hypoxemia): Those with hypoventilation and/or hypoxemia may have more going on than just OSA. An autoPAP may not restore oxygen to normal or correct shallow breathing. An in-lab PAP titration can test CPAP, and if ineffective, other more advanced devices can be tried like bilevel PAP (BPAP or BiPAP®), supplemental oxygen, or ventilator-like advanced breathing machines like volume-assured pressure support (VAPS). Patients need careful monitoring of their body position, breathing, oxygen, CO2, and their response to the PAP device being used to optimize the settings before PAP can be prescribed for home use.
Now that you are familiar with most of the reasons to have an overnight sleep study, here are a few reasons why you shouldn’t:
“It’s been five years since your last study, you need another study.”: FALSE. Time since the last study has nothing to do with needing another study, at least medically. Perhaps the register didn’t ka-ching for five years and is getting rusty. Or perhaps the cash registers have moved to tap and pay or contactless devices so there will never ever be a ka-ching ever again, now just a bleep. It still does not mean you need another study. Granted, there are some insurers that will deny patients payment for a new CPAP device or supplies if their sleep study is too “old,” but this is more likely a strategy to discourage utilization of resources, just adding more barriers to care. If you had OSA five years ago, unless you lost a lot of weight, you still have OSA. In fact, OSA gets gradually worse with age (and menopause!) so the five years is not helping the situation. Speaking of weight…
A little weight loss: Losing a little weight could tip someone with borderline OSA to less prominent OSA and under the arbitrary 5 breathing episodes/hr cutoff (called the AHI or REI), but will 5.5/hr to 4.2/hr move the needle in terms of the need for treatment or improvement of symptoms? Those with more moderate to severe OSA cannot expect resolution of the condition if they were to lose 10 or 20 pounds. Once you hit 50 or 100 pounds, now you’re talking. Unfortunately, in many with more severe OSA who went on to have bariatric surgery, many still had OSA after reaching their target weights. Pre-weight loss severity seems to correspond with improvement. Thus, if you lose a little weight, it is probably not worth repeating your sleep study. If you had severe OSA before significant weight loss, it is probably not worth getting another study either.
Insomnia: I see so many patients with insomnia and no risk factors for OSA getting sleep studies. Insomnia disorder is a chronic difficulty falling asleep, maintaining sleep, or getting back to sleep. Unless you have really severe and untreated OSA, your insomnia is not due to OSA and thus a sleep study is not helpful in understanding your sleep dfificutlies. Nothing like wearing dozens of electrodes and tubes and getting woken up at 5 am, while someone is watching you on a camera, in your nightgown, to help with the insomnia!!! OK, ok, the PSG is really not that bad, but the point here is why go through the inconvenience if it has nothing to do with your symptoms? First-line treatment is cognitive and behavioral therapy (CBT-I) as described in this video. It’s very hard to find. It is too bad insurance does not adequately reimburse sleep centers to administer CBT-I rather than paying a few thousand for a questionable sleep study.
Restless legs syndrome (RLS): Nowhere in the diagnostic criteria for RLS does it mention a procedure, much less an overnight sleep study to diagnose this neurological condition. How many RLS patients have I seen who have had a sleep study? Nearly all. The majority don’t have sleep apnea by the way. One could argue that periodic limb movements in sleep (PLMS) can be detected by a PSG. True. However, not all people with RLS have PLMS and the majority of those with PLMS do not have RLS. They are neither sensitive nor specific for the diagnosis of RLS. They are neither necessary nor sufficient to have RLS. If you did not catch the message amidst those sentence theatrics, detecting PLMS=not particularly helpful.
Overnight sleep studies are a challenge in the field of sleep medicine. Due to the insurance-based reimbursement system, they provide the revenue to sleep clinics when simply high-quality care for RLS, insomnia, or sleepy people will not. You can read why the direct care model is superior to this insurance-based approach for most sleep conditions. Unfortunately, if insurance stop paying for PSGs, most sleep centers would go under, and many have already with insurance requiring more home sleep apnea testing. There is the hint of some used car salesperson when an individual is told they need an overnight sleep study. The PSG is absolutely not physically harmful or all that invasive, maybe some sticky goo in your hair. Nevertheless, it still should be done for appropriate medical reasons only, as addressed in this blog. Aside from the inconvenience and hassle, there can be negative financial ramifications of doing a PSG, particularly for those paying cash, not meeting their high deductible, or holding a health-sharing plan among others.
One of my final patients when working for an insurance-based clinic had a borderline HSAT and then was told she needed an overnight study to test CPAP. She was billed $5000 and paid 20% co-insurance as she did it in a Tier 2 network. Her borderline HSAT was too mild to meet criteria for insurance to pay for CPAP. She paid $1000 out-of-pocket for the study, and she could not even get a CPAP. The cost of an auto CPAP? Less than $1000. It is important to educate yourself as a patient given the infiltration of the business interests of medicine into medical decision-making. Hopefully, some of the information above will help you decide whether you really do need an overnight sleep study. It is best only to do it if that’s where the money is for your own health.
-Andy Berkowski, MD of ReLACS Health, who has had at least four PSGs and more than a dozen HSATs, many of which were medically unnecessary but definitely for fun (and a little knowledge!)
