Self-Applied ECG Patch Catches Atrial Fibrillation More Quickly than Usual Care

Steven R. Steinhubl, MD

Through digital enrollment and utilization of participant-generated data from a wearable ECG patch to monitor for atrial fibrillation (AFib) can increase, markedly, the diagnosis rate relative to routine care—as much as 9-fold in the short-term and 3-fold in the long-term.

The findings of the mSToPS trial revealed that monitoring with a wearable ECG was additionally associated with a greater initiation of guideline-recommended therapy, although it did cause a greater health care resource utilization by year 1.

“At 1 year, people who wore the chest monitor had nearly 3 times the likelihood of being diagnosed with AFib, and appropriately. A significantly higher proportion of these patients were started on anticoagulant therapy to lower their stroke risk compared with those who received usual care,” Steven R. Steinhubl, MD, the director of digital medicine at Scripps Translational Science Institute, La Jolla, California and lead author of the study, said. “The data also provide a first glimpse into how this type of ECG screening might influence heart care utilization.”

Presented at the 67^th American College of Cardiology Scientific Sessions in Orlando, Florida, the trial sought to digitally enroll patients aged 75 and older with a prior cerebrovascular accident, heart failure, diabetes and hypertension, or obstructive sleep apnea (OSA), through either email (n = 52,553) or direct mail (n = 50,000). Ultimately, 2,655 consented and were randomized to either immediate monitoring (n = 1364) or randomized/delayed monitoring (n = 1291).

“Only 1.7% of eligible patients are enrolled in clinical trials, and less than a third of randomized clinical trials meet their original recruitment targets,” Steinhubl said. “But 88% of US adults use the internet and 77% own a smartphone, so in the context of a digital clinical trial, we sought to determine if participant-generated data, available through a wearable ECG patch, can improve the identification of AFib relative to routine care.”

The only statistically significant difference in baseline demographics between the immediate and delayed groups was the incidence of OSA, which was present in 24.9% of the immediate group and 29.0% in the delayed group (P = .02). Of the immediate monitoring group, 456 (33.4%) patients never wore an ECG patch, while 908 (66.5%) were actively monitored. In the delayed group, 457 (35.3%) never wore the patch, while 834 (64.6%) were actively monitored after 4 months.

Additionally, the groups were matched with 5310 observational controls to receive usual care, 3476 of which were available at the 12-month follow-up. At month 4, both the immediate and delayed groups were assessed for newly diagnosed AFib, with 0.6% in the delayed monitoring group having been diagnosed by month 4, and 0.2% in the immediate group, although diagnosis came much sooner (>20 days).

“The primary endpoint of the time to first diagnosis of AFib at year 1,” Steinhubl said. “AFib was defined as a reading of more than 30 consecutive seconds of atrial fibrillation by the ECG adjudicated by a clinical event committee, or a new diagnosis through claims data of a single new ICD9 or ICD10 code.”

At 1 year, AFib was newly diagnosed in 6.3% of those wearing a patch (unadjusted odds ratio [UnadjOR], 2.8; 95% CI, 2.1—3.7; P <.0001) compared to 2.3% in the observational group (adjOR, 3.0; 95% CI, 2.2—4.0; P <.0001). The active monitoring also resulted in a 2.0% increased intiation of anticoagulant therapy (5.4% vs. 3.4%; P =.0004).

The average patch wear time was 11.7 days in the groups wearing the ECGs, with median time until first AFib detection being 2 days (interquartile range, 1—5), bringing the median total AFib burden during monitoring to 0.9%. The median duration of the longest AFib episode was 185.5 minutes, with the majority of episodes—92.8%&mdash;occurring for more than 5 minutes, and 37.7% occurring for more than 6 hours.

None of the measurements for clinical outcomes between the actively monitored cohort and controls were revealed to be statistically significant, with rates of stroke (1.9% vs. 2.0%), myocardial infarction (1.78% vs. 1.84%) and systemic thromboembolism (0% vs. 0.03%) being essentially identical.

The study’s limitations included a lack of greater clarity for the clinical significance of “short” AFib episodes, as well as a lack of built-in digital prompting, which the authors believe is the reason that 38% of the consenting individuals never wore the patch. Additionally, although only 5.4% of individuals enrolled via email, Steinhubl and colleagues believe this may be due to the older population.

Currently, the lifetime risk of AFib for adults age 55 and older is 37%, although once AFib is present, it increases the risk of stroke 5-fold.

“A stroke is a very devastating cardiovascular complication of AFib; it can be tragic for patients and costly for payers,” Steinhubl said. “If we identify and get these people on a blood thinner, we can reduce their risk of stroke by about 70 percent, which is powerful.”

Click here to sign up for more MD Magazine content and updates.

Related Coverage >>>

iFR More Cost-Effective Than FFR for Patients Requiring PCI

STEMI Protocol Reduces Disparity Between Sexes, Improves Outcomes

Real-Time Genetic Data Impacts Physician Prescribing Behavior Following Stent Procedures