ADVERTISEMENT
Digital Health in the Electrophysiology Practice
In recent years, we have witnessed tremendous technological advancements not only in our daily life, but also in health care delivery. Cardiac electrophysiology (EP) is among the subspecialities that may benefit greatly from the potential of these new technologies. Wearable devices, biosensors, remote monitoring systems, and machine learning algorithms for the management of cardiac rhythm disturbances have already become a part of electrophysiologists’ routine practice. In this article, we provide a brief review of digital health technologies and their impact on the field of EP.
A New Paradigm Shift in Health Care
Traditionally, health care delivery and the emergence of new drugs or devices have always been physician directed. In this model, new advances in the management of patients were developed through the collaboration of physicians, researchers, and pharmaceutical companies, and were tested preferably with randomized clinical trials before reaching the patient. Today, giant tech companies such as Apple, Huawei, Amazon, and Google invest billions of dollars for the digitalization of the health care system. This is leading to a potential paradigm shift in health care from a physician- to a patient-directed model.
In outpatient EP clinics, more patients are presenting with cardiac rhythm disturbance alerts detected through their smartphones or smartwatches. In the majority of cases, these arrhythmia episodes were not previously diagnosed through days of electrocardiogram (ECG) recording or Holter monitoring. This has led to off-label use of medical devices before being tested in clinical trials or knowing their sensitivity or specificity in detecting rhythm disturbances. As more tech devices reach patients (ie, “consumers”) in coming years, EP societies, physicians, and tech companies share the important responsibility of regulating this new form of health care during the transformation phase.
The Road to Precision Medicine: Wearable Biosensors
Long-term data and analysis collected with ambulatory monitoring devices provide new options for diagnosis and management. Wearable electronic devices and biosensors have become increasingly popular for the continuous monitoring of heart disease. Most wearables track heart rate using photoplethysmography (PPG), an optical technique that can be used to detect changes in blood volume in the microvascular bed of tissue. These biosensors have been combined with algorithms to assess pulse irregularities and even diagnose atrial fibrillation (AF), which can be confirmed by a single-electrode ECG. While not all devices contain this feature, it is possible to obtain ECG data through wearable technology such as the Apple Watch (Apple). PPGs, which are flexible and can thus be placed on the finger, earlobe, forehead, or even the chest, can provide long-term and continuous monitoring in a noninvasive manner. In addition to passive PPG recordings, evaluation of contact-free facial and fingertip PPG measurements using a smartphone has demonstrated potential for screening and diagnosis of AF.1 These possibilities can speed up the process of diagnosing disease or detecting an abnormality, as well as improving patient compliance. However, considerations should be made for the large volume of data obtained from these devices as well as the instantaneous massive data transfer.
As wearable sensors become more reliable and affordable, they are expected to be widely used in the healthy population as well. The widespread use of wearable biosensors will be important when defining new AF burden thresholds and revising AF management strategies accordingly.
Use of Telemedicine and the Virtual Clinic (VC)
As a result of the COVID-19 pandemic, telemedicine and VC applications such as video conferencing have become more commonplace. By reducing the physical barriers to health care access, the need for personal protective equipment is removed and access to patients in isolation is facilitated. Advances in digital technologies and telecommunication also allow for more practical and cost-effective health care services for patients. Cardiac implantable electronic devices with wireless remote monitoring capabilities enable patient data to be acquired beyond the walls of the health care facility and transmitted back to the health care provider.2
A recent study demonstrated that VCs were effective in safely monitoring children with pacemakers or implantable cardioverter-defibrillators.3 Another study showed that virtual arrhythmia clinics were cost- and time-effective for the follow-up of AF patients undergoing catheter ablation.4 As a result, any potential issues could be detected earlier than with traditional outpatient follow-up.
In the future, digital health and VCs will undoubtedly play a larger role in the prevention, diagnosis, treatment, follow-up, and monitoring of chronic disease. With expanded use and application, a significant increase in the number of VCs is expected, further changing the way health care services are delivered.
Artificial Intelligence (AI) Algorithms
In recent years, we have also seen the rise of AI and machine learning (ML) applications in health care. The advancement of AI in mobile and wearable devices has introduced new strategies for disease detection, diagnosis, and follow-up.
Although the ECG has been in use for over 100 years, we are only discovering its full potential with the introduction of AI. The application of AI has transformed the ECG into a screening tool, enabling the prediction of cardiac and noncardiac disease. AI-enabled ECG algorithms are now being used to estimate patient sex and age, reveal the presence of asymptomatic AF during normal sinus rhythm, and identify patients with left ventricle systolic dysfunction.5-7 Additionally, a model developed with ML algorithms to predict clinical outcomes after cardiac resynchronization therapy implantation was shown to better differentiate outcomes over existing clinical discriminators.8 Applications such as those mentioned here show the potential of AI in clinical practice.
Application of AI to the ECG may also improve screening for hyperkalemia, as it was shown that subtle potassium changes could be detected using a deep learning model.9 With this application, it may be possible to titrate outpatients or change dialysis programs for drugs that impair potassium levels or kidney function. It could also allow for a bloodless test for serum electrolyte concentrations.
The application of ML methodologies to intracardiac data for better characterization of arrhythmia and determination of optimal ablation strategy needs further exploration. As the potential for AI expands, a framework for incorporating AI into clinical practice will also need to be developed. There is a need for applications that will seamlessly integrate AI models into clinical practice, taking into account data access and patient privacy.
Disclosures: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors have no conflicts of interest to report regarding the content herein.
References
1. Yan BP, Lai WHS, Chan CKY, et al. Contact-free screening of atrial fibrillation by a smartphone using facial pulsatile photoplethysmographic signals. J Am Heart Assoc. 2018;7(8):e008585. doi:10.1161/JAHA.118.008585
2. Slotwiner D, Varma N, Akar JG, et al. HRS Expert Consensus Statement on remote interrogation and monitoring for cardiovascular implantable electronic devices. Heart Rhythm. 2015;12(7):e69-e100. doi:10.1016/j.hrthm.2015.05.008
3. Spentzou G, Mayne K, Fulton H, McLeod K. Virtual clinics for follow-up of pacemakers and implantable cardioverter defibrillators in children. Cardiol Young. 2019;29(10):1243-1247. doi:10.1017/S1047951119001823
4. Manimaran M, Das D, Martinez P, et al. The impact of virtual arrhythmia clinics following catheter ablation for atrial fibrillation. Eur Heart J Qual Care Clin Outcomes. 2019;5(3):272-273. doi:10.1093/ehjqcco/qcz011
5. Attia ZI, Friedman PA, Noseworthy PA, et al. Age and sex estimation using artificial intelligence from standard 12-lead ECGs. Circ Arrhythm Electrophysiol. 2019;12(9):e007284. doi:10.1161/CIRCEP.119.007284
6. Attia ZI, Noseworthy PA, Lopez-Jimenez F, et al. An artificial intelligence-enabled ECG algorithm for the identification of patients with atrial fibrillation during sinus rhythm: a retrospective analysis of outcome prediction. Lancet. 2019;394(10201):861-867. doi:10.1016/S0140-6736(19)31721-0
7. Adedinsewo D, Carter RE, Attia Z, et al. Artificial intelligence-enabled ECG algorithm to identify patients with left ventricular systolic dysfunction presenting to the emergency department with dyspnea. Circ Arrhythm Electrophysiol. 2020;13(8):e008437. doi:10.1161/CIRCEP.120.008437
8. Kalscheur MM, Kipp RT, Tattersall MC, et al. Machine learning algorithm predicts cardiac resynchronization therapy outcomes: lessons from the COMPANION trial. Circ Arrhythm Electrophysiol. 2018;11(1):e005499. doi:10.1161/CIRCEP.117.005499
9. Galloway CD, Valys AV, Shreibati JB, et al. Development and validation of a deep-learning model to screen for hyperkalemia from the electrocardiogram. JAMA Cardiol. 2019;4(5):428-436. doi:10.1001/jamacardio.2019.0640
