National Heart, Lung, and Blood Advisory Council September 2019 Meeting Summary

DEPARTMENT OF HEALTH AND HUMAN SERVICES NATIONAL INSTITUTES OF HEALTH NATIONAL HEART, LUNG, AND BLOOD ADVISORY COUNCIL

MEETING SUMMARY OF THE NATIONAL HEART, LUNG, AND BLOOD ADVISORY COUNCIL

September 10, 2019

The 284th meeting of the National Heart, Lung, and Blood Advisory Council (NHLBAC) was convened on Tuesday, September 10, 2019 in Building 35A, the Porter Neuroscience Center Conference Center, National Institutes of Health (NIH), Bethesda, Maryland. In addition to NHLBAC members, the meeting included an ad hoc Board of Extramural Experts, scientists with research expertise in NHLBI mission areas who were recruited for this meeting as a special Council working group. The Council meeting began at 8:00 AM and concluded at 5:00 PM. Two working groups met from 10 AM to 11:15 AM and 2:30 PM to 3:45 PM. Dr. Gary H. Gibbons, Director of the National Heart, Lung, and Blood Institute (NHLBI), presided as Chair.

Council Members attending

Dr. E. Dale Abel
Dr. Donna K. Arnett
Dr. Jennifer Devoe
Dr. Serpil C. Erzurum
Dr. Karen Glanz
Dr. Garth Graham
Dr. Monica Kraft
Dr. Mohandas Narla
Dr. Diane J. Nugent
Dr. Robert C. Robbins
Dr. Kim M. Smith-Whitley
Dr. Kevin Thomas
Dr. Sally E. Wenzel
Dr. Andrew S. Weyrich

Board of Extramural Experts

Dr. Michelle Albert
Dr. Judy Aschner
Dr. Timothy Blackwell
Dr. Nadia Hansel
Dr. Bertha Hidalgo
Dr. Darrell Kotton
Dr. Brian Mittman
Dr. Matthias Nahrendorf
Dr. Ellis Neufeld
Dr. Laura Newby
Dr. Bruce Psaty
Dr. Susan Redline
Dr. Celeste Simon
Dr. Herman Taylor Jr.
Dr. Griffin M. Weber

Public attending

Dr. Xiaoping Bao, Purdue University
Dr. Uzay Emir, Purdue University
Dr. Nadia Hansel, Johns Hopkins University
Ms. Sheila Harley, BETAH Associates
Dr. Perry Kirkham, Purdue University
Mr. John Laughner, American Heart Association
Ms. Nuala Moor The American Thoracic Society

Other NIH Institute employees attending

A number of NHLBI staff members were in attendance.

I. CALL TO ORDER AND OPENING REMARKS

Dr. Gary H. Gibbons, Director of the National Heart, Lung, and Blood Institute (NHLBI), called the 284th meeting of the National Heart, Lung, and Blood Advisory Council (NHLBAC) to order and welcomed members and other attendees.

Dr. Gibbons noted that the 2016 NHLBI Strategic Vision is a culmination of a dynamic and collective effort in developing Strategic Research priorities that support four mission-oriented goals. The goals aim to understand human biology, reduce human disease, advance translational research, and develop the workforce and resources. The NHLBI is currently in the implementation phase of the Strategic Vision, which includes exploring how the research priorities (described in the NHLBI Strategic Vision) are being met, and how these priorities can be further addressed. The NHLBI Strategic Vision is intended to serve as a living document and this annual opportunity for joint feedback and discussion with the NHLBAC enables NHLBI’s Strategic Research Priorities to remain evergreen.

The meeting objective was to engage the NHLBAC as representatives of the NHLBI research community to advise the NHLBI on opportunities to advance and build upon the Strategic Vision in two rapidly evolving and cross-cutting areas: 1) Advancing precision prevention by modifying behaviors and affecting social determinants of health and 2) Advancing precision medicine by leveraging artificial intelligence and machine learning.

This meeting supported these objectives by providing two dynamic and thought-provoking speakers to inspire a robust dialogue. The first speaker, David Asch, M.D., M.B.A., addressed behavioral economics and new approaches to health care innovation. The second speaker, Eric Topol, M.D., spoke about personalized medicine and the role of artificial intelligence. Following each speaker, NHLBAC members broke into working groups to address key questions and develop bold cross-cutting ideas to further enhance research in these emerging areas with respect to the NHLBI mission.

II. ADMINISTRATIVE ANNOUNCEMENTS

Dr. Laura K. Moen, Director, Division of Extramural Research Activities, NHLBI, made the required announcements for the Council meeting, including the publication of a notice in the Federal Register as well as reminders to Council members regarding conflict of interest and lobbying activities.

III. ADVANCING PRECISION PREVENTION BY MODIFYING BEHAVIOR AND AFFECTING SOCIAL DETERMINANTS OF HEALTH

Dr. David Asch opened his talk by noting that we are in an era of discovery as evidenced by developments such as CRISPR, CAR-T, transcatheter aortic valve replacements. He pointed out that despite the fact that health care is much better now than it was in the past, people still lament the lack of innovation in health care. Dr. Asch posited that many do not see the discovery because it is the lived experience (long waits to see a physician or get an appointment, the cost of healthcare) that matters; it is the customer interface that is the reality. He gave examples of other popular businesses that have been successful because of their customer interfaces: Amazon, Airbnb, Uber. Dr. Asch also mentioned that we have been better at supporting breakthroughs in discovery than we have been at supporting those in productivity. He noted that health care is different than the examples he provided: it is high stakes, highly regulated, and capital intensive. Human behavior is irrational, and additional information is unlikely to alter behavior. However, health behavioral economics can facilitate future solutions because it can be designed to recognize the predictable irrationality of human behavior. Also, there are still concepts to be learned from other business models. Dr. Asch gave examples of some solutions that had been tried at his own institution and by others. He pointed out that humans model our behaviors on the behaviors of others; the status of the unwitnessed things made visible can be a powerful incentive; consider the ice bucket challenge for amyotrophic lateral sclerosis, stickers that identify someone as a blood donor. A question one might ask is how we make other’s activities witnessable in order to use them as an incentive? In closing, Dr. Asch noted that NIH has a role to be more innovative too.

IV. CHARGE OR WORKING GROUPS

Following Session 1, there were two breakout groups to discuss key questions. Their charge: Discuss novel, innovative, and cross-cutting means by which NHLBI can support and catalyze research to modify behaviors of patients, providers, and health systems to address social determinants of health.

V. REPORTS FROM WORKING GROUP DISCUSSIONS

The reports summarized below are ideas that were discussed by Council and BEE members and will be taken under advisement by NHLBI staff in future decision making.

Dr. Garth Graham and Dr. Annetine C. Gelijns summarized the discussions of Working Group 1A and Working Group 1B respectively by answering key questions. A summary of their input is described below:

• Data Integration – combine data on social determinants of health with clinical, genomic, omics to facilitate exploration drivers and potential interventions.
• Human Behavior – expand understanding of variation (heatmap) to refine interventions
• Population Heterogeneity – consider for personalized and community-based approaches/interventions
• Study Section/Review Process – support flexibility needed for integration of scientific disciplines to facilitate research on social determinants of health
• Expand Analyses – improve predictive modeling
• Health Systems Research – pilot trials and incorporate business models

Dr. Gibbons thanked the Council for their hard work and thought-provoking discussion. He added that NHLBI staff will review the feedback offered at this meeting and will use it in future planning and development.

VI. ADVANCING PRECISON MEDICINE BY LEVERAGING ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING

Dr. Eric Topol began by noting that a better term for precision medicine is individualized medicine. He framed his presentation in the context of errors and their ramifications; individualized medicine could potentially reduce the errors. Polygenic risk scores have been a big advancement in science with a potential to significantly affect clinical care. However, deep phenotyping involves many layers; Dr. Topol’s talk focused on sensors and the microbiome. One example of leveraging current technology: it is possible to get images for ultrasound with a probe connected to a smart phone. There is also a credit-card size 6-lead EKG sensor. Another example is the prevalence of blood pressure monitors that can be used outside of the clinic; there are even blood pressure watches that can be worn for continuous monitoring. If the use of these tools were widely accepted, it might be possible to reduce the cost and increase health care access if it can also reduce the number of visits to a clinic. Question is how useful is this information? There is a need for commercial recommendations; is it ethical to charge people for information when we don’t know its predictive capability? Deep phenotyping is what really shows who is at risk. Dr. Topol also demonstrated the limitations of the tools and machines. They pick up everything in an image; there is no discrimination about what may or may not be necessary for a diagnosis. He noted that these are still early days for the use of AI. There are few centers in the world using whole-slide digital imaging. There is a need for multimodal input, more than images are needed to bridge the gap between w here we are now and where AI could contribute in the future. Dr. Topol closed by pointing out that AI has tremendous potential to benefit those in health care professions as well as those who use the health care system. For example, if using AI resulted in one minute less time needed to input data per patient, the efficiency obtained in the health-care system could result in less burnout, and more time for empathy, human touch, and care. Questions from Council brought out the need for: (1) better datasets to represent all populations, (2) investment in infrastructure to handle the data and make it useable, and (3) plans for training of the health-care workforce.

VII. CHARGE TO WORKING GROUPS

• Charge: Discuss how NHLBI can advance precision medicine using artificial intelligence (AI) and machine learning (ML) to catalyze research and leverage/expand ongoing investments in data science.

VIII. REPORTS FROM WORKING GROUP DISCUSSIONS

The reports summarized below are ideas that were discussed by Council members and will be taken under advisement by NHLBI staff in future decision making.

Dr. Monica Kraft and Dr. Herman Taylor summarized the discussions of Working Group 2A and Working Group 2B respectively by answering key questions.

• Interface AI/ML – integrate with heart, lung, blood and sleep disorders at study inception; provide basic science insights.
• Simulations & Modeling – conduct with AI/ML then validate in clinical trials
• Generalizability & Bias – create networks to improve models, enhance heterogeneity
• Prospective evaluation trials – evaluate new technology on outcomes and care delivery
• Data Accessibility – connect electronic health records, widen accessibility, add underutilized clinical data, leverage cohorts.

Dr. Gibbons again thanked the Council members for their hard work and thought-provoking discussion. He added that NHLBI staff will review the feedback offered at this meeting and will use it in future planning and development.

COMMON THEMES FOR BOTH WORKING GROUPS

Both working groups saw a role for:

• Public-Academic-Private Partnerships – to promote innovation, learning, infrastructure
• Multi-disciplinary Research – to support and train teams with behavioral/data science experts
• Training Programs – leverage T32s and embed skills development in funding opportunities

ADJOURNMENT

The meeting was adjourned at 5:00 PM.