Apple Heart Study

Written by: Em Wessling, MD (NUEM ‘22) Edited by: Dana Loke, MD (NUEM ‘19) Expert Commentary by: Rod Passman, MD

Written by: Em Wessling, MD (NUEM ‘22) Edited by: Dana Loke, MD (NUEM ‘19) Expert Commentary by: Rod Passman, MD


Chief Complaint: My watch thinks I have Atrial Fibrillation!

 As technology advances, medicine must continue to advance in pace. Wearable technology has been evolving for decades. The information gathered from a wide range of these devices may someday help to provide healthcare workers with valuable information about a patient’s condition. However, for now, there is limited research on their utility within the healthcare field.

 Thus far, both Apple Watch and Fitbit have been shown to correctly identify tachycardia during atrial tachyarrhythmias, but their accuracy to the heart rate varied with the type of arrhythmia (1,2). Apple Watch has been shown to be more accurate than Fitbit (1,2). The WATCH AF trial demonstrated it was possible with reasonable sensitivity (93.7%) and specificity (98.2) to use smart watches to diagnose Atrial Fibrillation (3). 

How Apple Watch is tracking atrial fibrillation:

-        Photoplethysmography: the use of light to determine volume within a structure at a given time

-        Pulse is estimated by time between peak volume seen by photoplethysmography.

-        When the pulse is highly variable between consecutive beats, irregular heart beat is suspected.

 

Apple Heart Study: The plan and the preliminary data (4, 5)

Study Design: Prospective, single arm pragmatic study

-        Enrolled 419,093 participant  

-        Inclusion Criteria: appropriate Apple technology, Age≥22 years, US resident, proficient in English, valid phone number and email.

-        Exclusion criteria: self-reported atrial fibrillation , atrial flutter, or anticoagulation

-        Methods:  “Irregular Pulse Notification” (indication of possible atrial fibrillation) sent to participants  if 5/6 irregular pulses within  48 hour period, at which point participant was instructed to wear EKG patch for up to 7 days.

-        Primary Outcome: Proportion of patients alerted with “Irregular Pulse Notification” who were found to have atrial fibrillation or atrial flutter on EKG patch, in the 65+ population as well as in all-comers.

-        Secondary Outcomes: Positive predictive value (PPV) of irregular heart rhythm notification; percentage of those with irregular notification who contacted a health care professional within 3 months.

Preliminary Data presented at ACC:

-        Participants who received “Irregular Pulse Notification”:  2,161 (0.52% all comers)

-        Participants age >65 who received “Irregular Pulse Notification”: >3%

-        EKG Patches sent to 658 participants; 450 returned.

  • 34% of those returned showed atrial fibrillation

  • PPV for Tachogram: 71%

  • PPV for “Irregular Pulse Notification”: 84%

-        Notification to doctor - approx. 50%

Limitations:

  • Small sample size for EKG patches, despite high enrollment

  • Self-reported data

  • Self-selecting group, i.e.may not be able to extrapolate prevalence data to those who do not wear smart watches

 

Potential Impact on Emergency Departments:

As more and more studies validate the accuracy of wearable technology to measure and recognize health conditions, the implications must be analyzed as well.

Prior to 2017, researchers began to predict that there would be an expansive increase in the rates of atrial fibrillation due to “worldwide aging” (6). While this review acknowledged there were “potential applications” for smart phone technology in the diagnosis, their predictions of the expanse of the epidemic of atrial fibrillation preceded definitive research showing increased diagnosis rates with wearable technology, which will likely only further expedite this growing patient population.  The mSToPS Trial showed that immediate in-home monitoring with an EKG patch had 3% greater rates of atrial fibrillation diagnosis compared to delayed EKG monitoring at 4 months. This led to increased use of anticoagulants and increased health care utilization (7). If this increase was seen with EKG patches, consider the influx of patients to primary care and cardiology clinics in addition to emergency departments that can be projected based on the rise of smart watch detection of atrial fibrillation. Researchers in Australia had begun studying this prior to the commencement of the Apple Heart Study (8). When cardiac patients were asked if they trusted smart watches to predict arrhythmia and measure their heart rate only 53% agreed; however, that did not stop 91% from reporting they would seek care if their watch alerted them about an abnormality (8). While the preliminary data from the Apple Heart study shows that a much smaller percentage of those who were not previously cardiac patients sought medical care when alerted by their Apple Watch, further study is needed to see the extent to which advances in smart watch health technology will lead to an influx in patients to the Emergency Department due to concerns of arrhythmia found by a smartwatch (5).

 While the accuracy of these methods of arrhythmia detection are still being studied, the potential for ED presentation with this chief complaint will continue to rise. In the fourth quarter of 2017 financial year, Apple alone sold greater than 8 million smart watches worldwide, making it the largest watch vender in the world (9).  With these increased sales, comes the potential for increased recognition of arrythmia by smartwatch. Healthcare organizations throughout the country must strive to develop effective and efficient clinical pathways in order to evaluate, potentially diagnose, and treat this patient population. Upon presentation to the Emergency Department, each patient should receive an EKG, telemetry monitoring while in the Emergency Department and screaming lab work: often including CBC, BMP + Mg, and troponin.  From there, the pathway may vary. Many would agree, if the patient is and has always been asymptomatic, work up is unremarkable, with normal sinus rhythm on their EKG, discharge home with an EKG patch and follow up with cardiology is reasonable. Conversely, an EKG showing atrial fibrillation would constitute a new diagnosis and further work up would proceed as with any other new diagnosis of Atrial Fibrillation. However, for those who fall in-between, the disposition is not as clear. What would you do?


Expert Commentary

More than 800 years ago, Maimonides described an irregular pulse that likely represented atrial fibrillation (AF). The development of the electrocardiogram by Einthoven 700 years later allowed surface recordings of human AF for the first time.1 With the recognition that AF is often asymptomatic and paroxysmal, the development of inexpensive, non-invasive, passive monitors for irregular rhythm identification has long been recognized as a potentially important tool for arrhythmia detection and management.

At its core (pun intended), the purpose of the Apple Heart Study was to assess the feasibility of AF screening in large populations by monitoring participants with a wrist-worn photoplethysmography (PPG) monitor.2 The PPG algorithm in the Apple Watch samples the pulse several times daily during periods of physical inactivity and increases the sampling rate if an irregular tachogram is detected. If 5 out of 6 tachograms are consistent with AF (requiring > 60 minutes of AF), the wearer receives an irregular rhythm notification. Since the version of the Apple Watch used in the study did not have the 30-second ECG feature (available in Series 4 watches and later), the Apple Heart Study protocol asked those who received the irregular rhythm notification to wear an ECG patch at a later date. 

Several important facts can be gleaned from the Apple Heart Study.  First, the study virtually enrolled 419,297 individuals in less than a year, a testament to the interest in the subject matter, the ease of remote enrollment when appropriate, and the enormous potential of digital health studies. Second, the fear that the healthcare system would be inundated with false positive AF notifications appears unfounded as 99.8% of participants under age 40 did not receive an irregular rhythm notification.  Third, the positive predictive value for the irregular rhythm notification was surprisingly high (84%) despite that fact that the patch was applied a mean of 13 days following the notification and was worn for less than 7 days on average. This last point is worth emphasizing since with paroxysmal AF, a negative monitor placed two weeks after an irregular rhythm notification may simply mean that AF was not present during both time periods.

The study also has some important caveats.  The Apple Heart Study did not report the sensitivity and specificity of the PPG algorithm for AF detection, a critical piece of missing data needed for clinical care and future research. Furthermore, only a minority of patients who received an irregular rhythm notification actually wore and returned the ECG monitor, showing that virtual enrollment doesn’t always translate into virtual protocol compliance. From a research perspective, wearable AF monitors have allowed for large-scale screening studies such as the Huawei  Heart and Heartline Studies aimed at understanding the true prevalence of AF and the risks and benefits of early detection and treatment.3,4 From a clinical perspective, a patient who says “my watch says I have AF” still requires ECG confirmation, but that too has been made easier with the new generation of wearables.

References

1. Prystowsky EN. The history of atrial fibrillation: the last 100 years. J Cardiovasc Electrophysiol. 2008;19(6):575-582. doi:10.1111/j.1540-8167.2008.01184.

2. Perez MV, Mahaffey KW, Hedlin H, et al. Large-Scale Assessment of a Smartwatch to Identify Atrial Fibrillation. N Engl J Med. 2019;381(20):1909-1917. doi:10.1056/NEJMoa1901183

3. Guo Y, Wang H, Zhang H, et al. Mobile Photoplethysmographic Technology to Detect Atrial Fibrillation. J Am Coll Cardiol. 2019;74(19):2365-2375. doi:10.1016/j.jacc.2019.08.019

4. www.heartline.com

WbGJlUXI_400x400.jpg

Rod Passman, MD

Professor, Feinberg School of Medicine

Cardiac Electrophysiology

Northwestern Memorial Hospital


How To Cite This Post:

[Peer-Reviewed, Web Publication] Wessling, E. Loke, D. (2021, Jan 25). Apple Heart Study. [NUEM Blog. Expert Commentary by Passman, R]. Retrieved from http://www.nuemblog.com/apple-heart.


Other Posts You May Enjoy

References

1. Koshy, Anoop N., et al. "Smart watches for heart rate assessment in atrial arrhythmias." International journal of cardiology 266 (2018): 124-127.

2. Koshy, A., et al. "Heart Rate Assessment by Smart Watch: Utility or Futility?." Heart, Lung and Circulation 26 (2017): S280-S281.

3. Dörr, Marcus, et al. "The WATCH AF trial: SmartWATCHes for detection of atrial fibrillation." JACC: Clinical Electrophysiology5.2 (2019): 199-208.

4Turakhia, Mintu P., et al. "Rationale and design of a large-scale, app-based study to identify cardiac arrhythmias using a smartwatch: The Apple Heart Study." American heart journal207 (2019): 66-75.

5. ACC News Story. “Apple Heart Study Identifies AFib in Small Group of Apple Watch Wearers.” American College of Cardiology: Latest in Cardiology, American College of Cardiology, 16 Mar. 2019, www.acc.org/latest-in-cardiology/articles/2019/03/08/15/32/sat-9am-apple-heart-study-acc-2019.

6. Morillo CA, Banerjee A, Perel P, Wood D, Jouven X. Atrial fibrillation: the current epidemic. J Geriatr Cardiol. 2017;14(3):195–203. doi:10.11909/j.issn.1671-5411.2017.03.011

7. Steinhubl SR, Waalen J, Edwards AM, et al. Effect of a Home-Based Wearable Continuous ECG Monitoring Patch on Detection of Undiagnosed Atrial Fibrillation: The mSToPS Randomized Clinical Trial. JAMA.2018;320(2):146–155.

8. Koshy, A., et al. "Cardiac Patients Likely to Seek Medical Assistance Based on Abnormal Heart Rate Readings on Smart Watches or Smartphone ECG Monitors." Heart, Lung and Circulation 26 (2017): S280.

9. Canalys Press Team. “18 Million Apple Watches Ship in 2017, up 54% on 2016.” Canalys Newsroom, Canalys, 6 Feb. 2018, www.canalys.com/newsroom/18-million-apple-watches-ship-2017-54-2016.

Posted on January 25, 2021 and filed under Cardiovascular.