The first recording of cardiac electrical signals was done in 1887 by Augustus Weller, who observed changing cardiac electrical potentials with a mercury capillary electrometer.¹ A few years later, in 1901, Einthoven, rightfully referred to as the father of electrocardiography, developed the first string galvanometer electrocardiograph to record cardiac electrical signals, similar to a modern electrocardiogram (ECG).¹,² Although this device was relatively simple, none of the modern sophisticated electrocardiographs surpass the original Einthoven instrument in recording accuracy.² Einthoven’s immobile equipment was first used clinically to diagnose arrhythmias in

1906. This process required transtelephonic transmission of the ECG from the physiology laboratory to the clinic at the Academic Hospital of the University of Leiden, Netherlands, about a mile away, as described in the 1906 paper on the “télécardiogramme.”¹ Since then, the ECG has become an indispensable tool for the diagnosis and management of a wide variety of cardiac diseases, ranging from myocardial ischemia to cardiac arrhythmias and structural diseases of the myocardium, owing to its simplicity, availability, noninvasiveness, low cost, and reproducibility.³ Although a 12-lead ECG interpreted by a cardiologist has been traditionally considered the gold standard for arrhythmia diagnosis, in the last 5 years, single-lead, handheld ECG devices and smartphone-based ECG monitors with a built-in algorithm to differentiate between normal ECG and arrhythmias have revolutionized ECG recording and made it more accessible to the patient. As such, anyone who has a smartphone can become a user, and this may exponentially increase uptake of this technology. More than a hundred years from Einthoven’s first ECG recording, in the current era of telemedicine, the use of smartphones to record and transmit ECG data opens new horizons for arrhythmia diagnosis and monitoring.

A variety of handheld ECG devices are available, all of which produce diagnostic quality lead I single-lead ECGs (Figure 16.1). Most of these devices are equipped with automated algorithms to diagnose atrial fibrillation (AF), which provides superior accuracy compared with pulse palpation (sensitivity and specificity > 90%).⁴,⁵,⁶ Although single-lead ECGs may not always show P waves, their advantages, including ease of use, accessibility, and acceptance by the patients, outweigh this limitation.⁷ The majority of these devices record an ECG for 30 seconds, and they have been widely used in AF screening programs.

The Zenicor EKG (Zenicor Medical Systems AB, Stockholm, Sweden), also known as a thumb ECG, is a handheld single-lead ECG device, with which patients record their ECG signal themselves by placing their thumbs on two electrodes for 30 seconds (Figure 16.1A). By pressing a button the ECG data are transferred via the mobile network to a central ECG database, which stores and processes the ECG recordings received. A Web-based service allows physicians to view and interpret the ECGs in real time, and no additional equipment is needed for the transfer of data. This device was successfully used in large population-based AF screening studies in Sweden⁸,⁹ and has been shown to have higher sensitivity for detection of asymptomatic AF compared with conventional 24-hour Holter recordings.⁴,¹⁰

The MyDiagnostick (MyDiagnostick Medical BV, Maastricht, The Netherlands) is a handheld ECG device in the shape of a stick (length 26 cm, diameter 2 cm) with metallic electrodes at each end (Figure 16.1B). It is primarily intended to differentiate AF from sinus rhythm based on the ECG recording. The MyDiagnostick works by holding the device with both hands for 60 seconds until the result is revealed.
While holding the device, it will flash on the rhythm of the detected heartbeat. After 1 minute, either a green or a red light appears, indicating normal rhythm or AF, respectively. No ECG is displayed to the user. AF is diagnosed using an algorithm, based on the irregularity of RR intervals, provided that the arrhythmia is present during at least 75% (45 seconds) of the 1-minute ECG recording.⁶ Up to 140 1-minute ECG strips can be stored on the device. When more than 140 recordings are made, only the non-AF ECGs are overwritten, unless only AF recordings are present in the device. Physicians can have access to time- and date-stamped stored ECGs when the device is connected to a Web portal through a Universal Serial Bus connection on an Internet-enabled computer. The rechargeable battery allows for more than 300 recordings on a single charge, which corresponds to two to three daily ECG recordings for at least 3 months. This device was shown to have an excellent performance for the detection of AF in the primary care setting compared with a 12-lead ECG (sensitivity 100%, specificity 96%) and led to a new AF diagnosis in 1.6% of screened patients.⁶

The Omron monitor (model HCG-801, Omron Healthcare Europe, the Netherlands) is another single-lead handheld ECG device, which records an ECG by placing one electrode on the bare chest wall, while the other electrode is held by the right index finger of the patient (Figure 16.1C). The device monitor records a 30-second, single-lead ECG tracing and displays a message indicating the presence of possible AF, using an algorithm based on RR interval irregularity.⁵ The ECG recording and text message can be saved for downloading and analysis at a later time. This device demonstrated a high sensitivity (98%) for detecting AF in an elderly primary care population, but lacked specificity (<80%).⁵ It may thus be useful for ruling out AF, but a 12-lead ECG is likely to be needed to confirm diagnosis if the result comes back positive for AF.

The Merlin ECG event recorder (Meditech Ltd, Budapest, Hungary) is yet another device for recording a single-lead ECG. It differs from the aforementioned devices
in that it resembles a watch placed on the patients’ wrist (Figure 16.1D). To record an ECG, patients must cover the electrode on the face of the device with the palm of their right hand for 30 seconds. The recording does not provide an automated analysis but can be saved to a computer for later downloading and analysis. This device has a programmable recording time, from 15 to 300 seconds, and a full storage capacity of 15 minutes. Recordings are stored in the software database. This device had a 94% sensitivity and a 90% specificity compared with a 12-lead ECG for detecting AF in an elderly primary care population.⁵

In the current era, smartphones are ubiquitous, being carried around by an estimated 2 to 3 billion people.¹¹ Innovation in technology has produced smartphone-based single-lead ECG devices that enhance the feasibility of arrhythmia diagnosis and monitoring. These devices, which consist of an external case plus a downloadable application, allow a user to generate an ECG rhythm strip and confirm a rhythm, such as AF from their phone.

The first FDA-cleared over-the-counter ECG monitor, which was compatible with an iPhone called ECG check (Cardiac Designs, Inc., Park City, Utah), received FDA clearance in January 2013. This device utilizes Bluetooth technology for communication between the device and the smartphone. To record an ECG, the users are required to download a free application and place their fingers on the sensors on the ECG device. When this is done, a rhythm strip is created, which is viewable on the screen, stored on the phone, and transmitted to the ECG Check Web Center. The results from the rhythm strip are then transmitted back to the user in seconds. This device is now compatible with both iPhone and Android phones and is available both as prescription and as over the counter.

The KardiaMobile case or card (AliveCor, Inc., San Francisco, California) consists of two metal electrodes attached to a case (previous versions) or standalone (current version) that communicate with the phone through ultrasound technology (see Figure 10.17B). The device received FDA 510k approval in 2012, was marketed as a prescription-only heart monitor in 2013, and then received over-the-counter approval in 2014, as a single-lead cardiac event monitor. When the user touches each of the metal electrodes with his right and left hands, respectively, a bipolar ECG lead I is recorded for a period of 30 seconds. Cardiac electrical activity is transmitted to the smartphone by frequency modulation of an ultrasound signal using a 19-kHz center frequency and a modulation index of 200 Hz/mV. The ultrasonic signal received by the smartphone microphone is digitized at 44.1 kHz, 24-bit resolution, followed by demodulation of the signal to digital ECG tracing (300 samples per second, 16-bit resolution) by the application in the smartphone. The ECG tracing can be viewed in real time and is also stored, being instantly transmitted by the phone to a secure server for later review.¹² The recorded ECGs are classified as normal, unreadable, or AF, on the basis of an algorithm that evaluates for the presence of P waves and RR irregularity, resulting in 98% sensitivity and 97% specificity.¹² Figure 16.2 illustrates representative rhythm strip recordings, classified by the device as either normal (Figure 16.2A) or AF (Figure 16.2B), as viewed on the phone when held by a patient.

Not surprisingly, this product was met with much enthusiasm and was studied in multiple settings. The accuracy of this device, when compared with the 12-lead ECG for basic QRS morphology and PR interval measurements, was initially reported by our group.¹¹ The same first-generation case was subsequently compared with a standard 12-lead ECG on 381 individuals, including 128 healthy young adults, 123 Division I athletes, and 130 ambulatory cardiology patients.¹³ The study indicated that both smartphone and standard ECGs detected atrial rate and rhythm, atrioventricular (AV) block, and QRS delay with equal accuracy. The absolute differences in ECG intervals were small, with the higher variation in QRS duration and QT interval observed in patients with a paced rhythm and patients in cardiology clinics. Sensitivities to diagnose ECG abnormalities using an electrophysiologist’s interpretation as the gold standard, ranged from 77.4% for AV block to 94.4% for AF; the respective specificities were 96.4% and 99.4%. Through a postparticipation survey, the vast majority of participants reported that the device was convenient and easy to use, and most predicted that they would use it monthly. Most also stated that they would be willing to share such information with their health care provider.¹³

The feasibility and applicability of the KardiaMobile case to record an ECG was also shown in the pediatric population. In the Smartphone Pediatric ElectrocARdiogram (SPEAR) trial, which included 35 children, the reported user satisfaction was high. Of the 238 tracings, 96% were of diagnostic quality. Supraventricular tachycardia, ventricular tachycardia, ectopic atrial tachycardia, and AF were all recorded.¹⁴