A study from The University of Hong Kong in a primary care setting showed that the AliveCor Heart Monitor and Microlife WatchBP Office AFIB demonstrated less than expected sensitivity. 2052 patients (mean age = 68 years old) with hypertension and/or diabetes were screened for atrial fibrillation using these two automated devices and their algorithms.
The reference diagnosis was made by two blinded and independent cardiologists, who interpreted the signal from the AliveCor’s single-lead ECG. If either device detected atrial fibrillation, a 12-lead ECG study was conducted to confirm.
24 of the patients were diagnosed with atrial fibrillation. The AliveCor ECG detected 16 of 24 of them (sensitivity = 67%; 95% CI = 44.7-84.4%) and produced 11 false negatives (specificity = 99.5%). The Microlife WatchBP detected 20 of 24 of them (sensitivity = 83%; 95% CI = 63-95%) and produced 27 false positives (98.7% specificity).
Atrial fibrillation is a condition of the heart that causes it to beat abnormally. It is estimated to affect 33.5 million people worldwide and can be present without symptoms in 3% – 35% of patients around the world. In normal rhythm, a signal generated by the atria of the heart moves smoothly from the top to bottom, and then back up.
When atrial fibrillation is present, the top of the heart instead beats rapidly and without coordination, causing stuttering and therefore inefficient blood flow. In symptomatic patients, it typically causes feelings of palpitations, increased heart rate, shortness of breath, and exercise intolerance. It can sometimes cause chest pains if the rapid heartbeats lead to ischemia.
Importantly, atrial fibrillation can cause stroke (4-5% of patients), as chaotic blood flow can create blood clots that travel to become lodged in blood vessels of the brain. It’s therefore important to diagnose atrial fibrillation early, so that preventative measures can be taken, reducing risk of stroke by up to 80%.
Given this, it’s slightly jarring that in this study, the FDA-approved AliveCor ECG accurately caught only 67% of patients with stroke and the WatchBP caught only 83%. Interestingly, these results were much lower than the previous validation studies, which reported catching 98% and 98.5% patients using AliveCor and 94.9% using WatchBP.
This raises questions of their validity and use in a primary care settings where general practitioners would benefit from a quick screening algorithm to deliver timely care. The algorithms may need further refinement, and/or the clinicians may need further training to use the devices more accurately.