The FDA has approved algorithms for a number of SmartWatches. My confidence in FDA post-covid isn't as high as before but government experts feel they have benefit to public. False negative much more potential harm with a person ignoring signs of arrhythmia and not seeking cardiology consult. False positve not necessarily bad if it gets a high-risk person to cardiologist.
A number of cardiologists from prestigious University Medical Centers (Stanford, Cleveland Clinic, Harvard, Mayo Clinic) have YouTube videos recommending SmartWatches for people at higher risk for A Fib. Only 2% of Americans have A Fib in any given year so not a high volume issue.
Published Scientific Analysis have demonstrated "non-12-lead ECG ... offered high diagnostic accuracies for AF."
"Our search resulted in 16 studies using either non-12-lead ECG or PPG for detecting AF, comprising 3217 participants and 7623 assessments. The pooled estimates of sensitivity, specificity, PLR, NLR, and diagnostic odds ratio for the detection of AF were 89.7% (95% CI 83.2%-93.9%), 95.7% (95% CI 92.0%-97.7%), 20.64 (95% CI 10.10-42.15), 0.11 (95% CI 0.06-0.19), and 224.75 (95% CI 70.10-720.56), respectively, for the automatic interpretation of non-12-lead ECG measurements."
I am not sure if this was single lead or 6-lead. That is great for screening and early detection. "Diagnostic Accuracy of Ambulatory Devices in Detecting Atrial Fibrillation: Systematic Review and Meta-analysis" Tien Yun Yang et al. JMIR Mhealth Uhealth. 2021.
https://pubmed.ncbi.nlm.nih.gov/33835039/
Not many papers published on SmartWatch accuracy even though Apple has been out almost 5 years. I suspect nobody wants to "validate" so Apple can benefit commercially. Apple and other SmartWatch brands will need to do their own validation or pay big money for researchers to do it.
The algorithm isn't that complicated. Only need to measure milliseconds between R wave of QRS complexes to find irregularity and detect absense of P wave. That's not rocket science. One-lead ECG is sufficient as all other leads will show same conduction issue from different perspective. In A Fib, the other leads don't add much additional information unless you want to detect NSTEMI or Tachybrady Syndrome associated with A Fib.
A Fib from slow heart rate becomes more common in people as they age. And is just as deadly as fast A Fib. Particularly in people needing pacemaker, #1 reason for implanting pacemaker.
A Fib found on sleep study is less worrisome as CPAP generally "cures" this. No need to start anticoagulants if good nighttime CPAP prevents A Fib.
Another interesting paper of UK patients with self-monitoring data: "Patient-generated cardiovascular data were described in 185/1373 (13.5%) clinic letters overall, with the proportion doubling following onset of the first Covid-19 lockdown in England, from 8.3% to 16.6% (p < 0.001). In 127/185 (69%) cases self-monitored data were found to: provide or facilitate cardiac diagnoses (34/127); assist management of previously diagnosed cardiac conditions (55/127); be deployed for cardiovascular prevention (16/127); or be recommended for heart rhythm evaluation (10/127). In 58/185 (31%) cases clinicians did not put the self-monitored data to any evident use and in 12/185 (6.5%) cases patient-generated data prompted an unnecessary referral."
"Patient-initiated cardiovascular monitoring with commercially available devices: How useful is it in a cardiology outpatient setting? Mixed methods, observational study" Christine A'Court et al. BMC Cardiovasc Disord. 2022.
https://pubmed.ncbi.nlm.nih.gov/36175861/