Webinar Recap: Balancing Clinical Accuracy and Simplicity in HST


As part of Nox Medical’s commitment to the advancement of sleep diagnostics through simplification, increased efficiency, and comfort for all patients, the company launched a three-part webinar series titled “Advanced Sleep Diagnostics Through Home Sleep Testing” on March 17, 2021.

For the first session, Nox Medical welcomed Dr. Erna Sif Arnardottir, a leading sleep researcher and Director of the University of Reykjavik Sleep Institute, to discuss the challenges of today’s sleep diagnostics and home sleep testing. Titled “Balancing Clinical Accuracy and Simplicity in Home Sleep Testing,” the session covered the need for telehealth solutions in the dawn of COVID-19 and how providers can use home sleep testing to their advantage while remaining cognizant of potential gaps in accuracy.

The Shift to Home Sleep Testing

Dr. Arnardottir started by explaining how and why the industry has bridged the gap between in-lab polysomnography (PSG) testing, considered the “gold standard,” and home sleep apnea testing (HSAT) in recent years.

“There is a reason why we’re talking more about home sleep testing now than we would have been a year ago. There’s a huge need for home sleep testing and for having simpler ways of doing diagnostics because we appear to be in a sleep apnea pandemic as well. Nearly 1 billion people worldwide have sleep apnea, and that’s only accounting for adults.”

Once patients reach the point of undergoing sleep diagnostic testing for Obstructive Sleep Apnea, it’s standard practice to pre-screen patients with a questionnaire. Dr. Arnardottir warns of simplifying the process to the extent that practitioners only collect objective data, rather than also gathering subjective data, such as daytime sleepiness, sleep hygiene, comorbidities, etc., that is just as important to the screening process.

“We need to be careful when we do sleep diagnostics that we don’t simplify it to such a manner that we only do an objective sleep study or a home sleep study and don’t actually talk to the patient and get the subjective patient data. In our oversimplification to try and work faster and to do things better, we also need to make sure that the questionnaires and the patient experiences are recorded. This can be done for telemedicine and conducted online.”

While discussing why the gold standard is still so ineffective and what can be thrown out without jeopardizing quality, Dr. Arnardottir shares Sleep Revolution’s ongoing efforts to move diagnostics from the hospital to home in order to combat challenges facing PSG studies, such as long waiting lists and uncomfortable in-lab sleep settings.

As an advancement in the industry, she also references Nox Medical’s Self Applied Somnography (SAS) System* which allows the patient to self-apply the diagnostic technology using an instructional video in the comfort of their home. To obtain optimal amounts of data, the patient wears the technology for three consecutive nights, rather than one.

Balancing Simplicity and Clinical Accuracy in Home Sleep Testing

With the shift in diagnostics from a hospital to home setting, at what point in home sleep testing are we sacrificing too much data for simplicity? And how accurate is the data that is gathered from home sleep testing? To answer these valid questions, Dr. Arnardottir compared PSGs to Type III and Type IV devices and explained their limitations.

Type III Sleep Diagnostic Devices

Type III devices are typically used for patients with high pretest probability of obstructive sleep apnea (OSA) and utilize the same sensors as a PSG. However, the technology doesn’t record sleep itself or assess hypopneas causing arousals only. When using these devices, it’s imperative to gather subjective patient data to know whether or not a Type III device is a good fit or not.

“We need to remember that with these types of devices (Type III), that we do not have the actual sleep information. So the index time for the apnea-hypopnea index can lead to both over and underestimating sleep since it doesn’t distinguish between the awake and asleep. For people with milder disease, this could have a huge impact and may instead require a PSG.”

Type IV Sleep Diagnostic Devices

Dr. Arnardottir sees a bigger issue for simplicity with Type IV devices that only use an oximeter or a Peripheral Arterial Tonometry (PAT) device. She references a meta-analysis on how PAT technology works to diagnose OSA. As of 2014, the analysis included 14 studies and over 900 subjects.

“In comparison to PSG, the AHI was quite high. Meaning, on average, if you use this technology, you can have pretty good diagnostic accuracy. However, there are a lot of potential confounders.”

She explained that although PAT technology shows good performance measured against PSG on average, it may not work well for an individual patient as it excludes large patient populations, such as people with hypertension, which is typically half of the sleep apnea patients coming in for a diagnosis. Furthermore, it may not work well for younger people, and there are issues with comorbidities, etc., resulting in the likelihood of either under or over diagnosing sleep apnea. Also, since these devices do not measure the actual sleep signal, they result in a black box analysis — meaning there is an input and an output but no data to evaluate in between the two, she adds.

“When we are not measuring the actual signal, breathing or brainwaves, and if it is a black box analysis, we should always be careful. We need to think about all of the potential limitations of these devices, and even wearables.”

Going Beyond the AHI

When validating such devices, Dr. Arnardottir explains the need to evaluate the temporal overlap of events with an epoch-by-epoch comparison of the device to a PSG to note its accuracy. She also states that in order to go beyond the apnea-hypopnea index (AHI), the industry must work towards engaging with more than just Type III or IV devices and strive for high diagnostic accuracy and reliability despite the diagnostic tool.

“So even if we have so much to do, and there are so many patients waiting, we shouldn’t move to such simplistic ways of doing things that we actually lose all of our diagnostic accuracy and good way of doing things.”

Dr. Arnardottir ends by referencing the ongoing efforts of the Sleep Revolution, a four-year horizon project among 37 centers in Europe and Australia, to revolutionize sleep diagnostics and home sleep testing.

The session wrapped with a Q&A with Dr. Arnardottir where she answered compelling questions, such as the following:

  • Do you feel like the link between diagnosis and treatment is clear?
  • How do you prioritize who should go into the lab and who shouldn’t?
  • How do you feel about the general trust of the sleep diagnostic industry with artificial intelligence (AI)?

Missed the webinar? Catch up on the full presentation below.

The next webinar session titled Artificial Intelligence in Sleep Medicine with Dr. Jon S. Agustsson, will be held on Wednesday, April 7, at 11:00 am EDT / 5:00 pm CET. Please register HERE

*The Self Applied Somnography (SAS) is for research purposes only.

Topic: Industry News