Smart Ring Outperforms Apple Watch for Atrial Fibrillation Detection

A multicenter UK study comparing the diagnostic performance of the Apple Watch and Sky Labs' CART Ring has found that the ring-shaped wearable device outperformed the smartwatch in detecting atrial fibrillation (AF) in patients with cardiovascular disease. The research, led by Timothy Betts, MD MBChB, associate professor in the Oxford University Department of Cardiovascular Medicine, Oxford, UK, was published in the journal Heart Rhythm 02.1

Among 483 participants aged 18 years or older undergoing simultaneous 12-lead and single-lead ECGs (SL-ECGs), the CART Ring demonstrated significantly higher sensitivity and specificity for automated AF detection compared with the Apple Watch—84.6% vs 69.1% sensitivity and 89.9% vs 72.6% specificity, respectively (P <.01 for both). Authors explained that the findings were based on a “worst-case” analysis in which unclassified SL-ECGs were treated as diagnostic failures.1

In fact, unclassified readings were a major source of performance disparity, according to the study. The Apple Watch produced unclassified ECGs in 20.1% of recordings, compared to just 1.9% for the CART Ring. When Betts et al excluded these in a lenient analysis, both devices showed improved performance, and the difference in sensitivity narrowed (CART Ring, 86.4%; Apple Watch, 84.8%). However, the authors stated that the high unclassified rate for the Apple Watch remained a critical limitation in real-world use.

Physician interpretation of SL-ECGs considerably improved diagnostic performance for both devices. Sensitivity for detecting AF reached 95.4% for the Apple Watch and 94.3% for the CART Ring, with comparable specificities (89.6% and 88.9%, respectively). However, performance declined for arrhythmias other than AF. Both devices showed poor detection of atrial flutter and atrial tachycardia, particularly when the ventricular rate was regular. Physician interpretation in these cases also remained limited, with reported sensitivities of 37.5% for the Apple Watch and 22.9% for the CART Ring.1

Wearable devices have become widely used for detection of AF, Betts and colleagues wrote. But they cite data that show most validation studies of the technologies have enrolled only participants with either sinus rhythm or AF,2 “potentially overestimating diagnostic performance by excluding other clinically relevant arrhythmias, which may be important sources of both false-positive and false-negative detections,” they continued.2

Their current findings suggest that "hybrid models combining automated detection with physician interpretation could maximize the diagnostic utility of wearable ECG technologies in real-world practice.” They also noted that, “The Apple Watch missed approximately 1 in 3 episodes of AF and a high number of unclassified SL-ECGs,” underscoring the device’s limitations as a standalone screening tool.1

Betts and colleagues also noted while the majority of SL-ECGs in the study were of high quality, clinician interpretation of other rhythms was “suboptimal.” Given that the “heavily filtered” nature of the SL-ECG obscures certain features critical to diagnose other arrhythmias (eg, atrial flutter/tachycardia and heart block), they stressed the need for cautious interpretation of findings, reiterating that any management decision should be corroborated by 12-lead ECG.

Median age of the 483-participant cohort was 66 years and approximately one-quarter (29%) were women. Participants, recruited from 3 tertiary hospital in the UK, were enrolled between December 2021 and December 2022. Inclusion required a diagnosis of cardiovascular disease documented in the electronic health record (more at right). Participants were randomly assigned to 1 of 2 investigation arms with group 1 undergoing SL-ECG recording with the CART Ring first, followed by the Apple Watch, and group 2 following the reverse order, Apple Watch recording first, followed by the CART Ring.1

The study authors acknowledged several limitations. The cohort was composed primarily of older adults with established cardiovascular disease, which may limit generalizability to younger or asymptomatic populations. Importantly, only version 1.0 of the Apple Watch’s AF detection algorithm was assessed; newer versions, which increase the heart rate limit from 120 to 150 beats/min and incorporate improved sensors, may perform differently. Additionally, all SL-ECGs were recorded under supervision, likely improving signal quality compared to real-world conditions. Finally, the limited number of ECG attempts may have inflated the proportion of unclassified readings.1

Overall, the findings reinforce the need for physician involvement in the interpretation of wearable ECG data and suggest that device-specific factors—including signal quality and algorithm design—substantially affect diagnostic utility in clinical settings. Nonetheless they remain optimistic, writing, "As wearable technologies evolve, machine learning and artificial intelligence will likely continue to further improve their diagnostic metrics, learning through repeated use and training on increasingly large datasets."

References
1. Briosa e Gala A, Sharp AJ, Schramm D, et al. Diagnostic performance of single-lead electrocardiograms from the Apple watch and CART ring for cardiac arrhythmias. Heart Rhythm O2. 2025:1–10. doi:10.1016/j.hroo.2025.03.019 
2. Linz D, Gawalko M, Betz K, et al. Atrial fibrillation: epidemiology, screening and digital health. Lancet Reg Health Eur. 2024;37:100786. doi:10.1016/j.lanepe.2023.100786