Leveraging Artificial Intelligence to Drive Adoption of Remote Cardiac Monitoring

By Improving Data Management and Drastically Reducing False Positives, Artificial Intelligence Tools Make Remote Monitoring a More Practical Option for Electrophysiologists

In 2015, the Heart Rhythm Society published a consensus statement that bestowed remote monitoring (RM) with a Class 1A recommendation.¹ Since then, RM devices have been shown to reduce hospitalizations^2-4 and the risk of death from heart failure (HF).^1,4

Yet, adoption of these care-enhancing solutions has been relatively slow. By 2020, only 1 in 5 hospitals and health clinics had adopted RM tools.⁵

Part of this delay is due, no doubt, to the impact of the COVID-19 pandemic. RM’s demonstrated utility during the pandemic helped maintain clinical connections between patients and providers without the need to be in the same physical location.⁶ During the international lockdown in early 2020, Iacopino et al observed that 81% of events were associated with a remote alert and 60% of RM events triggered a clinical action.⁷

Another factor may be inertia. It’s difficult for any industry, especially one as large and complex as health care, to adopt a new way of doing things. Antiquated health care reimbursement models and poor integration with electronic health record systems have made it difficult for providers to use RM at scale or get proper resource commitment from administration.      

In addition, the sheer volume of data produced by RM devices such as implantable loop recorders (ILRs) presents challenges to the everyday use of RM strategies.⁸ Providers do not have the tools to develop intuitive, time-saving workflows that present large volumes of disparate data in a unified and actionable manner.

To overcome these challenges, providers must leverage artificial intelligence (AI) tools that aggregate, analyze, and help manage these new streams of patient information. The main objective is to reduce the workload on health care professionals (HCPs) by applying filtering systems on alerts to reduce the number of false positive events.

The Challenges and Opportunities of Cardiac Monitoring With RM Devices

For decades, patient electrocardiograms (ECGs) were commonly monitored with surface 24-hour Holters or transtelephonic monitoring of cardiac implantable electronic devices (CIEDs), such as implantable cardioverter-defibrillators (ICDs), pacemakers, and cardiac resynchronization therapy (CRT) devices.

Today, millions are monitored and diagnosed via sophisticated CIEDs emitting electrogram signals, and long-term measurement of ECGs from subcutaneous ILRs. This shift has significantly improved HCPs’ ability to determine the significance of arrhythmias and provide timely therapeutic options.

Due to the need for improved diagnosis of arrhythmias such as atrial fibrillation (AF) (whether already diagnosed or suspected in the case of cryptogenic stroke), symptomatic ventricular tachycardia (VT), or asystole, ILRs are widely and increasingly used in routine care.⁹ They now represent a significant part of arrhythmia-related diagnosis, most notably for cryptogenic stroke and unexplained syncope, and to a lesser degree, palpitations and AF management.

However, despite improved device algorithms and increasing RM deployment, the actionable rate of alerts is significantly low. For example, the TRUST study demonstrated that only 6.6% of programmed transmissions of RM in patients with ICDs required clinical action.¹⁰ It has even been shown that only 3.2% of the transmissions per patient result in a clinical action.¹¹

This problem is particularly acute for patients with ILRs. False positives range consistently from 46% to 86% in recent studies, depending on indications.¹² While the devices need to be sensitive enough to detect virtually all possible instances of a given issue, the consequence is that many events considered arrhythmias are normal rhythms. These false positives aren’t mere annoyances. They burden HCPs with extra work and effectively reduce the number of patients that any given clinic can support through RM tools.

“COVID-19 has had a very special effect in the world of RM. We found there was a significant adoption of RM during the pandemic, and that is here to stay. That is continuous. People have learned from that experience and realized that this is a very efficient and effective way of managing patients,” said Niraj Varma, MD, PhD, from the Cleveland Clinic in Cleveland, Ohio.

“One of the discouragements was the amount of data that people had to handle and the number of patients that had to be managed with RM. As the volume of patients with implantable devices increases, the volume of data that they transmit increases. Managing this workload with restricted resources is a challenge. So we were moved to an actionable care model that still requires a lot of data management, our institution included. We have moved towards using some third parties, and Implicity is one of them, that can manage these data and within the framework of that technology, practice what we call alert-based care.”

Fortunately, emerging technologies such as AI can help providers balance sensitivity and specificity. Robust AI algorithms can filter out the signal from the noise, limiting HCPs’ exposure to only the most actionable data elements for each patient.

Reducing RM False Positives With AI

AI has found a home in health care, as many of the industry’s modern challenges involve enormous volumes of highly complex data. AI computational techniques are uniquely suited to identify patterns in health care data and surfacing actionable insights to clinicians without excessive manual investment.

As HCPs look to scale up their use of RM, AI tools will become a crucial support solution. By running data from RM systems through appropriate AI algorithms, clinicians can drastically reduce the number of false alarms and avoidable wellness checks, thus improving patient care and reducing the burden on clinical staff.

In December 2021, Implicity, a clinical algorithm company developing software for medical devices, announced FDA clearance for a novel AI algorithm that analyzes ECG data from ILRs. ILR ECG Analyzer^* tool is an AI-based medical algorithm specifically designed to flag and remove false positive episodes. ILR ECG Analyzer applies AI to the heart rhythm data collected from specified insertable cardiac monitors, improving the accuracy of irregular heartbeat detection and prioritizing “true” events that warrant further action.

New research demonstrates that Implicity’s AI tool can reduce the number of ILR incidents that need review by one-third; the algorithm reduced the false positivity rate by 79% during the study while maintaining 99% sensitivity.¹³ Of more than 2800 episodes processed by ILR ECG Analyzer, more than 1200 were reclassified as normal rhythm. In a clinical setting, each of these reclassified incidents would translate to a reduction in the resources required to provide appropriate care.

“Implicity’s ILR ECG Analyzer is a one-of-a-kind innovation that will allow for efficient, accurate, and much more timely adjudication of transmitted ILR events,” said Kevin Campbell, MD, FACC, from Health First Medical Group in Merritt Island, Florida. “ILR ECG Analyzer is the first FDA-approved tool utilizing AI to quickly identify those transmissions that require clinical action. I firmly believe that this technology will result in more timely interventions and improved outcomes for our patients.”

Charting a Path Toward the Future of Actionable Care Models for Cardiology Patients

As patients and providers adopt RM technologies, both have seen the benefits of this new strategy for accomplishing shared clinical goals.

Providers have become far more likely to detect a problem through the continuous use of RM. Even if a patient isn’t experiencing symptoms, RM can provide earlier detection so the patient can be brought in for care. In the future, we’re likely to see the rise of an “actionable care” model with reimbursement models to match. Patients will receive proactive, meaningful care when they need it, based on actionable alerts and detailed data for informed decision-making.

“I am excited by the potential this technology has to transform the practice of remote patient monitoring,” said Prof Varma. “The true promise of RM is that we can see patients who need to be seen earlier. Even if the patient is not experiencing a symptom, we can detect their condition and let them know they need to be seen. Implicity’s innovative solutions will enable us to direct our attention to clinically actionable data better so we can do just that.”

In addition to the clinical benefits, more medico-economic studies have observed the positive economic impact of RM. As the use of RM increased between 2010 and 2015, annual costs per beneficiary have decreased.^4,14,15 One review of ICDs and CRTs showed increased cost-effectiveness with RM implementation.¹⁵ As RM adoption continues, this effect is likely to be magnified.

The future is bright for RM in cardiology, with several emerging uses on the horizon. New possibilities for RM of CIEDs exist for the prediction and prevention of worsening HF. For example, using a weight scale connected to the Implicity platform, HCPs can follow the rapid weight gain potentially linked to the early decompensation of HF.

Transmitted data, even without an additional device connected when an implant is available, will soon allow early prediction of HF decompensation. Prediction and prevention are the 2 major arms of future cardiology care thanks to AI-based algorithms. They could significantly impact the health system’s cost through improved evolution of the disease by avoiding heavy treatments, reducing hospitalizations, and limiting deaths linked to cardiovascular diseases. Moreover, scaling this RM activity could also mean more time spent with patients when needed.

RM represents a vast improvement in the way care has traditionally been delivered. AI tools such as ILR ECG Analyzer will likely play a key role in helping HCPs manage and use RM data for immediate improvements in patient care, along with a platform that can aggregate, normalize, and standardize data from all CIEDs across all major manufacturers, with an ergonomic interface to improve the workflow.

“In the future, we’ll have an actionable care model. That means that patients will be continuously monitored,” predicts Prof Varma. “When the device notes a problem, it will immediately identify the patient and notify the clinic, and I think there are significant patient benefits to practicing this form of remote monitoring.” 

^*United States’ Food and Drug Administration (FDA)-cleared Class II medical device and Conformité Européenne (CE) marked Class I (under Medical Device Directive [MDD]) medical device. See instructions for use for more information.

To learn more about Implicity, visit https://www.implicity.com/.

This article was published with support from Implicity.

References

1. Saxon LA, Hayes DL, Gilliam FR, et al. Long-term outcome after ICD and CRT implantation and influence of remote device follow-up: the ALTITUDE survival study. Circulation. 2010;122(23):2359-2367. doi:10.1161/CIRCULATIONAHA.110.960633

2. Assaad M, Sarsam S, Naqvi A, Zughaib M. CardioMems® device implantation reduces repeat hospitalizations in heart failure patients: a single-center experience. JRSM Cardiovasc Dis. 2019;8:2048004019833290. doi:10.1177/2048004019833290

3. Tajstra M, Sokal A, Gadula-Gacek E, et al. Remote Supervision to Decrease Hospitalization Rate (RESULT) study in patients with implanted cardioverter-defibrillator. Europace. 2020;22(5):769-776. doi:10.1093/europace/euaa072

4. Chew DS, Zarrabi M, You I, et al. Clinical and economic outcomes associated with remote monitoring for cardiac implantable electronic devices: a population-based analysis. Can J Cardiol. 2022;38(6):736-744. doi:10.1016/j.cjca.2022.01.022

5. Holland TM. 6 remote patient monitoring lessons learned from COVID-19. Samsung. Published July 26, 2021. Accessed July 7, 2022. https://insights.samsung.com/2021/07/26/6-remote-patient-monitoring-lessons-learned-from-covid-19/

6. Magnocavallo M, Vetta G, Bernardini A, et al. Impact of COVID-19 pandemic on cardiac electronic device management and role of remote monitoring. Card Electrophysiol Clin. 2022;14(1):125-131. doi:10.1016/j.ccep.2021.10.010

7. Iacopino S, Placentino F, Colella J, et al. Remote monitoring of cardiac implantable devices during COVID-19 outbreak: “keep people safe” and “focus only on health care needs. Acta Cardiol. 2021;76(2):158-161. doi:10.1080/00015385.2020.1847459

8. O’Shea, Middeldorp ME, Hendriks JM, et al. Remote monitoring of implantable loop recorders: false-positive alert episode burden. Circ Arrhythm Electrophysiol. 2021;14(11):e009635. doi:10.1161/CIRCEP.121.009635

9. Giancaterino S, Lupercio F, Nishimura M, Hsu JC. Current and future use of insertable cardiac monitors. JACC Clin Electrophysiol. 2018;4(11):1383-1396. doi:10.1016/j.jacep.2018.06.001

10. Varma N, Epstein AE, Irimpen A, Schweikert R, Love C. Efficacy and safety of automatic remote monitoring for implantable cardioverter-defibrillator follow-up: the Lumos-T Safely Reduces Routine Office Device Follow-up (TRUST) trial. Circulation. 2010;122(4):325-332. doi:10.1161/CIRCULATIONAHA.110.937409

11. Ninni S, Delahaye C, Klein C, et al. A report on the impact of remote monitoring in patients with S-ICD: insights from a prospective registry. Pacing Clin Electrophysiol. 2019;42(3):349-355. doi:10.1111/pace.13598

12. Afzal MR, Mease J, Koppert T, et al. Incidence of false-positive transmissions during remote rhythm monitoring with implantable loop recorders. Heart Rhythm. 2020;17(1):75-80. doi:10.1016/j.hrthm.2019.07.015

13. Rosier A, Crespin E, Lazarus A, et al. B-PO04-037: A novel proprietary algorithm reduces the false positive rate of Medtronic LNQ11 ICM devices by 79%. Heart Rhythm. 2021;18(8):S294. https://doi.org/10.1016/j.hrthm.2021.06.733

14. Ricci RP, Vicentini A, D’Onofrio A, et al. Economic analysis of remote monitoring of cardiac implantable electronic devices: results of the Health Economics Evaluation Registry for Remote Follow-up (TARIFF) study. Heart Rhythm. 2017;14(1):50-57. doi:10.1016/j.hrthm.2016.09.008

15. Sequeira S, Jarvis CI, Benchouche A, Seymour J, Tadmouri A. Cost-effectiveness of remote monitoring of implantable cardioverter-defibrillators in France: a meta-analysis and an integrated economic model derived from randomized controlled trials. Europace. 2020;22(7):1071-1082. doi:10.1093/europace/euaa082

Suggested Reading

1. Crossley GH, Boyle A, Vitense H, Chang Y, Mead R, CONNECT Investigators. The CONNECT (Clinical Evaluation of Remote Notification to Reduce Time to Clinical Decision) trial: the value of wireless remote monitoring with automatic clinician alerts. J Am Coll Cardiol. 2011;57(10):1181-1189.

2. Mabo P, Victor F, Bazin P, et al, COMPAS Trial Investigators. A randomized trial of long-term remote monitoring of pacemaker recipients (the COMPAS trial). Eur Heart J. 2012;33(9):1105-1111. doi:10.1093/eurheartj/ehr419

3. Simovic S, Providencia R, Barra S, et al. The use of remote monitoring of cardiac implantable devices during the COVID-19 pandemic: an EHRA physician survey. Europace. 2021;33(9):1105-1111. doi:10.1093/eurheartj/ehr419

4. 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):e69-100. doi:10.1016/j.hrthm.2015.05.008