IEEE Future Directions
Related Articles
Just two years ago I was at a conference discussing the value of wearable as medical devices. The general mood among practitioners was that their accuracy was too low and too dependent on the way they were used to take the data generated into any consideration.
The situation, however, seems to be different today.
A study, just published by Stanford -on March 16th, 2019-, reports on results obtained by having 400,000 people across 50 US states using the Apple watch for over a year, since November 2017. They were required to wear an Apple watch (series 1, 2, 3 – series 4 was not available at the time of the study) and an iPhone.
Over this period Stanford researchers detected notifications of potential atrial fibrillation to just 0.5% of users (that is around 2,000 notifications). This relatively low number is important because you do not want to cry wolf and then discover there is nothing to cry about. Do that a few time and when there will be a reason no one will take notice.
Actually, the study found that 84% of those who received a notification did have an atrial fibrillation at the time of the warning. Notable, 34% of those who received a notification of irregular pulse and who took an ECG a week later were found to have atrial fibrillation. This is a significant number given that atrial fibrillation is an intermittent condition that may not be manifest if you take an ECG after some time.
It should also be noted that only 57% of those who received an alert notification sought medical advice. It is not 100% but it is still a significant number. It shows that people are not used to receive these types of alert and downplay their relevance. Over time it is likely that trust on these devices will increase and with it an appropriate response to those notification will be taken,
It is not just about atrial fibrillation. Wearables have the potential to detect a number of other situations requiring medical attention. Most of these are intermittent in nature and difficult to detect in a periodical screening. Having the possibility to be alerted and to show the notification to your doctor may trigger more in depth analyses and eventually save lives.
The lower accuracy of wearable (as I mentioned in most cases the lower accuracy is not an issue of the device, rather the result of its improper use) is counterbalanced by the extensive time the wearable has to monitor us. Over this time spikes and errors can be evened out and intermittent situations may be detected. Accuracy will result from the many measures taken, rather than from the accuracy of a single measure.
Apple is clearly betting on the value of wearable and this study is supporting that.
In the next decade I am pretty sure that having a monitoring device on us, most of the time, will become normal, and I expect the avalanche of data that will become available will trigger the development of many applications that will change the way we look at healthcare.
Just read this news on the web:
“Posted to Reddit by “ClockworkWVVII” with the title “My Apple Watch just saved my life,” the Redditor explains they received a high heart rate notification while relaxing in bed. The user then contacted the emergency services for medical assistance, as a preventative measure.
The paramedics arrived to find the user in “serious trouble,” with their body described as having gone “into shock.” The poster was put onto a stretcher and taken to the trauma center of the local hospital. ”
https://appleinsider.com/articles/19/04/08/apple-watchs-heart-rate-monitoring-credited-with-saving-another-life?utm_source=Daily+Email&utm_campaign=bca9826b18-EMAIL_CAMPAIGN_2019_04_08_02_42&utm_medium=email&utm_term=0_03a4a88021-bca9826b18-248625753