‘A powerful tool’
Overall, cardiac electrophysiologists showed high agreement in differentiating between AFib and non-AFib, with high interobserver reproducibility. A manual diagnosis was not possible for 10% of tracings because of either poor ECG quality (3%) or unclear P-waves (7%).
Fifty-nine of the 580 patients in SR were misclassified as AFib by the experts, and 5 of the 154 patients in AFib were misclassified as SR.
“Our results show that the presence of sinus node dysfunction, second- or third-degree AV block, ventricular paced rhythm, PVCs, and IVCD were more frequently represented in smartwatch misdiagnoses,” wrote the authors. “Patients with PVCs were three times as likely to have false positive AFib diagnoses.”
Study limitations included the single-center nature of the study and the fact that patients were recruited in a cardiology office. The latter factor may have influenced the incidence of ECG abnormalities, which was much higher than for the average smartwatch user.
“Even with its limitations, the smartwatch remains a powerful tool that is able to detect AFib and multiple other abnormalities,” said Dr. Strik. “Missed diagnosis of AFib may be less important in real life because of repeated measurements, and algorithms will continue to improve.”
Technology improving
Richard C. Becker, MD, director and physician in chief of the University of Cincinnati Heart, Lung, and Vascular Institute, said, “This is exactly the kind of investigation required to improve upon existing detection algorithms that will someday facilitate routine use in patient care. An ability to detect AFib in a large proportion of those with the heart rhythm abnormality is encouraging.”
The findings should not detract from well-conducted studies in otherwise healthy individuals of varied age in whom AFib was accurately detected, he added. “Similarly, an automatic diagnosis algorithm for AF, pending optimization and validation in a large and diverse cohort, should be viewed as a communication tool between patients and health care providers.”
Patients at risk for developing AFib could benefit from continuous monitoring using a smartwatch, said Dr. Becker. “Pre-existing heart rhythm abnormalities must be taken into consideration. Optimal utilization of emerging technology to include wearables requires an understanding of performance and limitations. It is best undertaken in coordination with a health care provider.”
Andrés F. Miranda-Arboleda, MD, and Adrian Baranchuk, MD, of Kingston Health Sciences Center, Canada, conclude in an accompanying editorial, “In a certain manner, the smartwatch algorithms for the detection of AFib in patients with cardiovascular conditions are not yet smart enough ... but they may soon be.”
The study was supported by the French government. Dr. Strik, Dr. Miranda-Arboleda, Dr. Baranchuk, and Dr. Becker reported no conflicts of interest.
A version of this article first appeared on Medscape.com.