Wrist-Worn Wearable Reflectance Pulse Oximeters During Sleep: How Accurate are They?
We want to know how accurate a wrist-worn wearable device with a built-in oxygen saturation monitor is when you’re sleeping. Does it accurately detect your oxygen saturation levels, giving more insight into possible conditions? Let’s explore:?
What are wearable reflectance pulse oximeters??
Wearables have come out with many new features in the last few years that are helpful in detecting and diagnosing certain health conditions. One being wearable reflectance pulse oximeters (wrist-worn oxygen saturation monitors.) These pulse oximeters work by using two light sources and a photodiode,1 which is the sensor that receives the light reflected onto the skin and converts this light into an electrical signal.2 These light sources are also able to detect blood pressure and heart rate. This study outlines how wrist-worn reflectance pulse oximeters are “safe, convenient, inexpensive, and non-invasive” and are clinically accepted alternatives for oxygen saturation monitoring.1??
What is polysomnography (PSG)??
Polysomnography (PSG), or a sleep study that detects sleep patterns in order to identify sleep disorders and their causes 3 has historically been the gold standard in detecting and determining the severity of sleep apnea (the apnea-hypopnea index, or AHI) and other sleep-related breathing disorders.4? However, it comes with its own set of limitations, “such as high costs, lengthy time requirements, and the need for an in-hospital setting.”4??
How accurate are wearables??
Because of PSG tests’ high costs and lack of accessibility, other methods of detecting sleep-related breathing disorders via wearable devices have become more popular.?
One study used the Samsung Galaxy Watch 4, which takes an oxygen reading every second, to detect SpO2 during sleep. The oxygen desaturation index acquired by the wearable was then compared to the AHI during sleep and used pulse oximetry on the fingertip as a point of reference. The accuracy of wrist-worn pulse oximeters can be negatively impacted by too much movement, and so “SpO2 values were screened using the morphology of PPG waveforms and stability of the output data.”4 This caused the watch to reject any data with distorted waveforms and excessive movement.?
Out of 97 adults tested, 672.3 sleep hours were recorded with 3.4 hours excluded from fingertip pulse oximetry and 171.1 hours from the wrist SpO2, due unreliable data. Four categories of sleep apnea were looked at: normal, mild, moderate, and severe.?
During the study, it was found that the watch was imprecise in patients with very severe OSA but did generally well in the other three groups. 73.5% of the data from the wearable pulse oximeter was viable for use, with a rejection rate of 26.5%.4?
Study Results?
Overall, it was found that when SpO2 decreased during sleep, the wrist pulse oximeter had a lower value than that of the reference. The study writers suggest that with improvements to accuracy and error correction algorithms, future devices will be more reliable and accurate for clinical use.4??
领英推荐
Conclusion of Wrist-Worn Pulse Oximeters?
Although wrist-worn pulse oximeters are convenient and generally more cost-effective compared to full sleep studies or even fingertip pulse oximeters, the accuracy also must be taken into consideration. During the last study mentioned, which is (to the best of our knowledge) the only study done evaluating the accuracy of wrist-worn reflectance pulse oximetry during sleep, as the patient’s oxygen saturation decreased, the accuracy of the wrist-worn pulse oximeter also decreased. Wrist-worn pulse oximeters are more accurate during periods of decreased movement. As stated above, wrist-worn pulse oximeters worn during sleep are not the most accurate for detecting sleep-related breathing disorders and should have accuracy and error correction algorithms updated before being used in the clinical setting.?
?
Cardiogram’s Detection of Sleep Apnea?
Even though the use of a wearable pulse oximeter has not found to be the most accurate during sleep, Cardiogram has a feature for sleep apnea worth looking into. In this study , Cardiogram was used with a deep neural network to help detect sleep apnea, hypertension and diabetes using only heart rate and step count. That deep neural network, DeepHeart, “is a semi-supervised deep neural network that accurately predicts cardiovascular risk, but requires 10x less labeled data than conventional deep learning techniques.”5 During this study, DeepHeart was able to accurately detect hypertension 84.8% of the time, sleep apnea 90.4% of the time, and “results were not statistically significant for diabetes.”6??
Although some studies have found wrist-worn pulse oximeters to be less than accurate during sleep, certain technologies increase the effectiveness of these wearables, allowing us to accurately detect conditions.?
?
Sources?