We're in the midst of a fourth industrial revolution, fueled by a doubling of computer performance every two years for over a half-century that has provided the platform for technological breakthroughs including machine learning, artificial intelligence (AI), robotics, nanotechnologies and quantum computing. The associated exponential increase in digitization has led to transformation and disruption across all sectors of the economy, including health care.

Simultaneously, we're facing the transition to value-based care and the ongoing need to promote population health, spurring a growing interest to adopt innovative health delivery models that incorporate data from wearable devices and remote digital sensors. The Centers for Medicare and Medicaid Services also is promoting the use of remote monitoring of patients, which will increase its inclusion in home health delivery. Additionally, the Food and Drug Administration has provided guidance on software as a medical device, a requisite component for signal processing of the prodigious amounts of data potentially produced by remote devices.

Evolving Data Tsunami: Call to Action

Clinicians can prepare for the data tsunami resulting from these technology and regulatory changes by addressing two key questions:

• How can we strategically leverage real-time data coupled with AI tools to augment our clinical workflow in this digitized world?
• How can we make the path to analytics-based prediction and prescription smooth and more efficient, without incurring data fatigue and digital burn out?

Opportunities For Cardiovascular Clinicians

Wearable devices and remote digital sensors already provide a plethora of physiological, clinical and disease marker data. Early action by the clinician on these data could make it possible to avoid, or least lessen the impact of, adverse and critical health events. What's needed to do this? Real-time data processing and analysis in a time-sensitive manner via signal processing systems.

Use Cases in Cardiovascular Medicine

Real-time data analysis should be a great asset to clinicians addressing cardiovascular health in three important use cases:

• Heart failure. For remote monitoring of patients with heart failure, various biosensors and implantable hemodynamic monitoring devices have been developed. Parameters including heart rate, blood pressure, weight, ECG, edema, pulmonary artery and right ventricular pressures can be monitored. Real-time analysis of this complex multidimensional data could markedly improve management and prevent disease exacerbation and readmission.
• Arrhythmia. For patients at risk for arrhythmia, remote arrhythmia monitoring devices and other types of wearable sensors are increasingly being used. Real-time analysis of these data can help to shorten the time to arrhythmia detection and treatment, thus avoiding major complications.
• Hypertension. For patients with poorly controlled hypertension, real-time analysis of remote blood pressure monitoring data to evaluate hidden patterns and trends would assist planning of personalized management and improve long-term outcomes.

Potential Solutions: Clinician's Workflow Augmentation

Solutions must augment clinician workflow by providing actionable analysis – a smart, real-time dashboard powered by analytics. In this model, the stream of data from remote digital sensors and wearable devices is processed by real-time health analytics, appropriate data mining and machine learning algorithms, and predictive and prescriptive analytics are facilitated.

The visual output is displayed on an electronic health record- (EHR) integrated clinical dashboard with high impact findings messaged directly to clinicians, providing-real time actionable insights. This dashboard output is concise, clinically relevant, pre-interpreted and straightforward for the clinician to understand and execute requisite action.

Advantages: Augmented Medical Intelligence

Innovative use of real-time analytics supported by machine learning technology applied to remote health delivery models will provide augmented medical intelligence and promote the quadruple aim in cardiovascular medicine through:

• Improved quality of care and patient satisfaction, by providing risk-stratified prompt, personalized treatment that will improve patient outcomes and promote wellness.
• Lower health care costs, by reducing response time to critical data and avoiding impending adverse health events, unnecessary hospital admissions, emergency department visits and complications due to undue progression of cardiovascular disease.
• Promotion of population health, by improved health care delivery to all patients with chronic heart diseases.
• Improved clinician experience, by avoiding data fatigue/burn out related to the massive load of continuously generated data from cardiovascular monitoring devices.

Challenges

Among the challenges of addressing the data tsunami are assuring proper routing of the data stream, data privacy, security and governance. Accuracy of predictive algorithms and models and interpretability of their output are critical for proper clinical decision-making and prevention of diagnostic and management errors. Smooth integration within the clinician workflow and appropriate remuneration are additional challenges.

Collaboration among regulators, payers, providers, data scientists and machine learning experts will be required to develop smart, real-time actionable clinical dashboards powered by real-time analytics and AI to assist the early detection of diseases and prevent adverse health events. Early intervention will lead to better quality of care, improved health outcomes and ultimately success in the era of value-based medicine.

---

This article was authored by Jai Nahar, MD; Arash Harzand, MD; Andrew M. Freeman, MD, FACC; Regina S. Druz, MD, FACC; and James E. Tcheng, MD, FACC.

Keywords: ACC Publications, Cardiology Magazine, United States Food and Drug Administration, Patient Satisfaction, Heart Rate, Early Intervention, Educational, Remuneration, Reading, Reaction Time, Burnout, Professional, Centers for Medicare and Medicaid Services, U.S., Blood Pressure, Patient Readmission, Pulmonary Artery, Ventricular Pressure, Medicaid, Medicare, Computers, Software, Arrhythmias, Cardiac, Data Mining, Artificial Intelligence, Electronic Health Records, Health Care Costs, Heart Failure, Disease Progression, Electrocardiography, Edema, Heart Diseases, Intelligence, Early Diagnosis, Nanotechnology, Cardiovascular Diseases, Biosensing Techniques, Publishing, Emergency Service, Hospital, Hypertension, Cognition

< Back to Listings