At-Home Sleep Recording Makes Strides Thanks to Wearables, AI

As the prevalence of sleep disorders such as obstructive sleep apnea has grown, so too has the need for sleep specialists. Yet, a new report suggests the trends may not necessarily mean a spike in the need for sleep laboratories.

In a new article in the journal Sleep Medicine Clinics, corresponding author Henri Korkalainen, Ph.D., and colleagues, explained how wearable devices, artificial intelligence and a growing demand for sleep testing are fueling a push toward at-home testing.

Henri Korkalainen

“There is certainly a rising interest in the field for home-based [sleep] recordings,” Korkalainen told Managed Healthcare Executive®. “Moreover, the overall interest in sleep and wellbeing has risen over the past years, and the awareness of sleep disorders, such as obstructive sleep apnea, has increased significantly.”

At present, the “gold standard” for sleep diagnostics is polysomnography conducted within a specialized sleep laboratory. Such testing includes a number of measurements, such as brain activity, eye movement, chin and leg muscle tone, cardiac function, respiratory effort, and blood oxygen saturation, among others. Such tests are also accompanied by video and audio recordings.

While all of that data can help clinicians zero in on an accurate diagnosis, it also comes with significant drawbacks.

“Overall, the sleep recordings conducted in a specialized sleep laboratory are expensive, and besides, they still force the patient to sleep in an unfamiliar environment under constant surveillance, and this can hinder normal sleep,” said Korkalaienen, a postdoctoral researcher at the University of Eastern Finland.

Sleep labs also can have long wait times to get an appointment, and patients are typically only observed for a single night.

There are also less onerous types of sleep monitoring. The Task Force of the Standards of Practice Committee of the American Sleep Disorder Association has come up with four types of sleep monitoring, including unattended monitoring that tracks fewer types of data. Those categories include at-home devices, though Korkalainen and colleagues said the classifications were made two decades ago. Since then, major improvements in sensors, data analysis, and recording technology have been made. He and colleagues said newer wearable devices have boosted the potential of at-home testing, to the point where even electroencephalography (EEG) can be done at home.

“Moreover, advancements in deep learning applications in the field of sleep medicine are especially helpful for home recordings, as they can enable automatic analysis and even provide good estimates of things such as sleep architecture from surrogate signals that can easily be measured,” he said.

One of the most valuable tools, he and colleagues argue, is a pulse oximeter. Though a relatively simple device, it can conduct photoplethysmograms, which, thanks to deep learning technology, can be used to identify sleep stages and match sleep scoring done with polysomnography with moderate agreement.

Korkalainen said at-home testing is not perfect. Recording quality may suffer in an at-home environment, and it is possible devices will fail to record at all.

“In order to mitigate these risks, clear and informative instructing of the patient is crucial and requires expertise from the healthcare professional,” he said.

When patients do home-testing, there is a possibility that they will need re-testing, and Korkalainen said that is an important factor to consider when analyzing the costs and benefits of at-home testing versus sleep-lab testing. However, he said existing research has shown that at-home testing is usually cheaper, even if it must be repeated.

Still, he said there are some limits to at-home testing.

“The home recordings are certainly already a viable option for diagnosing some sleep disorders (e.g. obstructive sleep apnea) and their importance will likely increase in the future,” he said; “however, they cannot replace in-lab studies in all cases, for example, when multiple comorbid sleep disorders are suspected.”