The ECG application, which is linked to the Apple Watch and is designed to detect signs of cardiac arrhythmia, generates relatively few false alarms. This is clear from the results of a study published in the journal English Journal of Medicine, conducted by researchers from Stanford University in California and a team from Apple. For the investigation, more than 400,000 security guards over eight months have agreed to evaluate their pulse data using Apple's algorithms that are not publicly available and telematically contacted for suspected atrial fibrillation.
The IT group already had its first figures from the Apple-funded study within application availability in countries like Germany in March. Now, experts can use the article to look at the entire design of the study and the details of the analysis. For example, the publication states that the manufacturer uses the algorithm to alert stakeholders if this makes five of the six received signals irregular.
Low rate of return
During the experiment, this was the case for a total of 2161 persons or 0.52 percent of participants. They received a mobile ECG sticker by mail that was supposed to take seven days to confirm suspected atrial fibrillation. 450 people returned the adhesive, of which 153 (34 percent) actually confirmed the first "diagnosis". Of those notified through the application, about half participated in a follow-up survey. 76 percent said they sought additional medical help following Smart Watch tips.
Overall, the study had unexpectedly low response rates, which calls into question the informational value of the data. In addition, most of the participants were socio-economically better and younger, and thus not representative of the population. These and other critical points have already attacked other scientists for comment. At the same time, they described the method of extensive investigation as a "pioneering step forward".