key: cord-1020421-1jxe58vo authors: Basza, Mikołaj; Krzowski, Bartosz; Balsam, Paweł; Grabowski, Marcin; Opolski, Grzegorz; Kołtowski, Łukasz title: An Apple Watch a day keeps the doctor away? date: 2021-12-31 journal: Cardiol J DOI: 10.5603/cj.2021.0140 sha: 7bee685adfdd9afe69615452ad92418bc5964c3a doc_id: 1020421 cord_uid: 1jxe58vo nan The high prevalence of cardiovascular diseases and an increasing health workforce shortage are major challenges that cardiology will face in the decades to come. New technologies supporting physicians in screening and patient monitoring are considered a potential solution to help fill this gap. For example, the average smartwatch user generates millions of basic vital sign data points each day, observed almost continuously 7 days a week in a patient's natural environment. 2019 ended with more than 500 million wearable users, and according to the market forecasts, this number will have reached over 1 billion by 2022 [1] . Apple dominates the smartwatch market with Apple Watch (AW), which is recognized as a Food and Drug Asscoiation class-2 medical device; nevertheless, its applications and reliability in medicine remain unclear. The AW collects various biosignals based on two core technologies, photoplethysmography (PPG) and single-lead electrocardiography (ECG) [2] . Using green or infrared light, PPG is used to measure heart rate (HR) and HR variability, relying on blood volume changes in tissue. According to manufacturer information, the AW supports an HR range of 30-210 bpm. Measurements of HR are taken in the background during rest, but a user can also perform it on-demand. In many studies with similar conclusions, HR measurement accuracy has been validated on a healthy population during different activities. Overall accuracy during rest, low and moderate activity is comparable to standard 12-lead ECG and decreases near range thresholds with motion artifacts [3] [4] [5] . These findings suggest a possible use of AW HR monitoring during therapy and cardiac rehabilitation. Apple offers an HR notification function that informs the user when the HR in a 10 min resting period is out of the previously set range. The irregular rhythm notification function is widely described in the Apple Heart Study [6] . AW showed a 34% diagnostic yield in detecting atrial fibrillation (AF), 35% in the population over 65 years old and 18% in the cohort younger than 40 years of age. 89% of true positives had AF episodes lasting at least 1 hour. The positive predictive value of individual notification was 84%. The high false-positive rate was the main allegation in this study. However, ECG patches used for validation were worn for an average of 6.3 days and a few days after initial notification and by a limited number of study participants, so the absence of AF during validation does not indicate that the notification was false. Especially in paroxysmal and infrequent episodes of arrhythmia. This study's findings imply the potential benefits of irregular pulse notification in AF screening in a population over 65 years old when AF screening is cost-effective. According to the European Society of Cardiology (ESC) 2020 guidelines, methods based on PPG require additional ECG confirmation, which is possible in Apple Watch Series 4 and higher and occasionally, in case of doubt, needs to be verified with traditional 12-lead ECG tracing (Fig. 1) . The Apple Watch Series 4 and higher have built-in two electrodes, a titanium electrode in the Digital Crown, and a chromium silicon carbon nitride layer applied on the back of the AW. They allow obtaining a single lead ECG corresponding to lead I in standard ECG. In a validation study, Apple showed high morphology agreement between Apple Watch ECG and 12-lead ECG (lead I) waveforms at rest and after exercise. Sub-study of the Leipzig Apple Heart Rhythm Study compared three leads obtained using alternative positioning of Apple Watch to leads I, II, III of standard 12lead ECG with the correlation between 97% and 99% for a duration (ms) of P wave, PR interval, QRS complex, QT interval and T wave in all three leads. The polarity concordance of P wave, QRS complex, and T wave was 98-100%. However, a lower correlation -72-90% -has been observed for those parameters' amplitude (mV). In 2020 Spaccarotella et al. [7] obtained leads I, II, III, V 1 , V 2 , V 3 , V 4 , V 5 , and V 6 using AW and compared it to 12-lead ECG on 100 patients admitted to the coronary care unit. Bland-Altmann did not show a significant difference in millimeters of ST-segment deviation. The feasibility of AW ECG interpretation in AF detection has been confirmed with 96% sensitivity and 100% specificity [8] . Those studies have shown that Apple Watch ECG waveform is a clinically relevant biosignal and meets the criteria of ESC 2020 guidelines for AF definitive diagnosis, which is especially important in patients at high risk of stroke. Single lead ECG is the basis of the Apple Watch ECG app 1.0 algorithm that, according to the manufacturer, is characterized by 98.3% sensitivity and 99.6% specificity in detecting AF. Still, this data requires cautious interpretation. An independent study conducted on 50 patients, who had undergone cardiac surgery, monitored by telemetry, showed 41% specificity in detecting AF by an algorithm and 5 times higher unreadable ECGs rate [8] . Apple Watch ECG app 2.0 algorithm interprets the rhythm automatically and classifies it into sinus rhythm, AF, low or high heart rate, inconclusive and poor recording. Apple's internal study demonstrated 98% correctly classified sinus rhythms and AF in the HR range of 50-99 bpm and 93% for sinus rhythm, and 83% for AF in HR range 100-150 bpm. There are no other studies that confirmed the results of the ECG 2.0 Clinical Validation Study. One of the AW ECG waveform's possible applications is screening and monitoring therapy with a cardiac adverse effect. In 2020 Strik et al. [9] validated QT interval measurements using AW ECG to assess potential usefulness in patients with COVID-19 on hydroxychloroquine/azithromycin treatment. He revealed that adequate QT measurements were possible from AW ECG in 85% of patients. However, using additionally lead II and V 6 obtained using AW increased that score to 94%. Year by year, the manufacturers of wearables are getting deeper into healthcare by providing new medical features and conducting further studies in collaboration with leading medical institutions. In Apple Watch Series 6, optical sensors were redesigned to provide oxygen saturation and VO 2 max measurements. Although the blood oxygen level app is officially not intended for medical use, it might be a medical-grade feature in the future. Apple recently has launched two new studies investigating the feasibility of the AW in cardiology. One is related to monitoring heart failure patients, and the second is focusing on the correlation between physical activity and heart health [10] . Though there is a need for further validation, studies have shown that wearables such as AW should not be viewed just like gadgets. Still, they may become a reliable source of medical data. In addition, professional guidance and clinical experience could provide enormous potential in screening arrhythmias, supporting therapy, and cardiac rehabilitation. Cisco Visual Networking Index (VNI) Global and Asia Pacific Mobile Data Traffic Forecast Using Apple watch for arrhythmia detection Validity and reliability of the Apple Watch for measuring heart rate during exercise Accuracy of Apple Watch measurements for heart rate and energy expenditure in patients with cardiovascular disease: cross-sectional study Accuracy of consumer wearable heart rate measurement during an ecologically valid 24-hour period: intraindividual validation study Large-Scale assessment of a smartwatch to identify atrial fibrillation Multichannel electrocardiograms obtained by a smartwatch for the diagnosis of ST-segment changes Accuracy of Apple Watch for detection of atrial fibrillation Validating qt-interval measurement using the Apple Watch ECG to enable remote monitoring during the COVID-19 pandemic Apple Heart & Movement Study (NCT04198194)