Mobile health technology to detect AF led to a high rate of false positives in patients with certain cardiac conditions.
Mobile health technologies to detect atrial fibrillation (AF) are linked to a high rate of false positives and inconclusive results in patients with certain cardiac conditions, according to new findings.
The largest real-world study to date indicated that the use of these devices is particularly challenging in patients with abnormal electrocardiograms (ECGs). Both improved algorithms and machine learning could help the tools provide more accurate diagnoses, according to investigators.
“With the growing use of smartwatches in medicine, it is important to know which medical conditions and ECG abnormalities could impact and alter the detection of AF by the smartwatch in order to optimize the care of our patients,” said lead study author Marc Strik, MD, PhD, LIRYC institute, Bordeaux University Hospital. “Smartwatch detection of AF has great potential, but it is more challenging in patients with pre-existing cardiac disease.”
In theory, the use of extended cardiac monitoring in patients and the use of implantable cardiovascular electronic devices may increase detection of AF. However, limitations with the devices include a short battery life and a lack of immediate feedback.
New smartphone tools that have the ability to record an ECG strip and make an automated diagnosis may thus overcome the above limitations and lead to a timely diagnosis. Strik noted that prior studies have validated the accuracy of the Apple Watch for the diagnosis of AF in a “limited number of patients with similar clinical profiles.”
The investigators performed a test on the accuracy of the Apple Watch ECG app in the detection of AF in patients with a variety of coexisting ECG abnormalities.
Their study included a total of 734 consecutive hospitalized patients. Each underwent a 12-lead ECG, with immediate follow-up by a 30-second Apple Watch recording.
Investigators reported that each smartwatch’s automated single-lead ECG AF detections were classified as “no signs of atrial fibrillation,” “atrial fibrillation,” or “inconclusive reading.” The recordings were given to an electrophysiologist who performed a blinded reading, assigning each tracing a diagnosis of “AF,” “absence of AF,” or “diagnosis unclear.” A second blinded electrophysiologist analyzed 100 randomly selected traces to determine “the extent to which the observers agreed.”
The findings indicated that the smartwatch ECG failed to produce an automatic diagnosis in approximately one in every five patients.
Additionally, investigators reported the risk of having a false positive automated AF detection was higher for patients with premature atrial and ventricular contracts (PACs/PVCs), sinus node dysfunction, and second- or third-degree atrioventricular-block.
For those in AF, the risk of having a false negative tracing (missed AF) was reported as higher for patients with ventricular conduction abnormalities (intraventricular conduction delay) or rhythms controlled by an implanted pacemaker. Moreover, the cardiac electrophysiologists had a high level of agreement for differentiation between AF and non-AF.
Data indicate the smartphone app correctly identified 78% of patients in AF and 81% who were not in AF. Meanwhile, the electrophysiologists identified 97% of patients who were in AF and 89% who were not in AF.
Those with PVCs were three times more likely to have false positive AF diagnoses from the smartwatch ECG according to the data, and the identification of patients with atrial tachycardia (AT) and atrial flutter (AFL) was considered very poor.
“These observations are not surprising, as smartwatch automated detection algorithms are based solely on cycle variability,” Strik added. “Ideally, an algorithm would better discriminate between PVCs and AF. Any algorithm limited to the analysis of cycle variability will have poor accuracy in detecting AT/AFL. Machine learning approaches may increase smartwatch AF detection accuracy in these patients.”
The article, “Role of Coexisting ECG Anomalies in the Accuracy of Smartwatch ECG Detection of Atrial Fibrillation,” was published in the Canadian Journal of Cardiology.