Summary of second annual MCBK public meeting: Mobilizing Computable Biomedical Knowledge—A movement to accelerate translation of knowledge into action

Abstract The volume of biomedical knowledge is growing exponentially and much of this knowledge is represented in computer executable formats, such as models, algorithms and programmatic code. There is a growing need to apply this knowledge to improve health in Learning Health Systems, health delivery organizations, and other settings. However, most organizations do not yet have the infrastructure required to consume and apply computable knowledge, and national policies and standards adoption are not sufficient to ensure that it is discoverable and used safely and fairly, nor is there widespread experience in the process of knowledge implementation as clinical decision support. The Mobilizing Computable Biomedical Knowledge (MCBK) community formed in 2016 to address these needs. This report summarizes the main outputs of the Second Annual MCBK public meeting, which was held at the National Institutes of Health on July 18‐19, 2019 and brought together over 150 participants from various domains to frame and address important dimensions for mobilizing CBK.


| BACKGROUND
Despite significant healthcare spending in the United States, our health outcomes are worse than countries that spend far less. 1 Morbidity and mortality data indicate growing health disparities, 2 despite the (overwhelming amount of) accumulated knowledge from biomedical research.
Undeniably, many factors cumulatively influence the health of our nation, but it appears that achieving the goal of routinely and equitably applying biomedical knowledge when and where it is needed remains elusive.
Biomedical knowledge is growing at a dramatic pace, 3 and there is a pressing need to incorporate or provide access to this knowledge into a variety of systems, organizations, and applications. Increasingly, biomedical knowledge is represented in computable formats with potential for rapid dissemination. [4][5][6] Computable biomedical knowledge (CBK), such as predictive models or executable rules, alerts, or data visualizations, can be integrated into health information systems to improve health care and the in-context learning of trainees. CBK is essential for both the discovery and intervention components of learning health system activities. 5 The movement to mobilize computable biomedical knowledge (MCBK) aims to represent knowledge in computable formats and make it widely available to achieve better health in different settings, contexts, and applications.
While there has been a great deal of research in knowledge sharing methods and techniques, only recently have more robust standards emerged for encoding and specifying knowledge in a computer interpretable form. This development, combined with the broad adoption of EHR technologies and the continuing pressures toward payment reform, have made the need for shareable computable knowledge paramount. The MCBK movement crystalized 3 years ago with a handful of clinical and informatics thought leaders who believed that the mobilization of CBK was essential to improving health and health care. In 2016, approximately 40 experts met at the University of Michigan to explore strategies to define requirements for infrastructure, standards, policies and best practices around CBK. 7 In 2018, the First Annual MCBK public meeting was held at the National Library of Medicine and drew over 140 participants. 8 Planners and participants developed a Manifesto (see Supporting Information) to articulate shared values and principles around CBK that would maximize the benefits, and minimize the harms, of widespread use of CBK for health, including imperatives to expose the underlying evidence and currency of CBK, and for efficient, safe, and equitable diffusion. 9 Achieving the ideals of the Manifesto suggests an ecosystem with many actors and processes for generating new knowledge, evaluating it, applying it, and monitoring its effects. The MCBK movement is committed to shaping the development of an open ecosystem that will make CBK easily findable, accessible, interoperable, and reusable (FAIR). 10

| MEETING AND PARTICIPANT INFORMATION
The Second Annual MCBK public meeting was held July [18][19]2019 in the Natcher Conference Center at the National institutes of Health   She spoke about models as the foundation of CBK and NLM's aspiration to advance science by supporting access to reproducible and reusable models. Dr. Brennan mentioned recent NIH guidance encouraging use of the Fast Healthcare Interoperability Resources (FHIR) standards specification for research applications and data sharing, 12 and suggested that the unprecedented momentum around FHIR could energize the MCBK community and enable the use of CBK when and where it is needed.

Dr. Don Rucker, National Coordinator for Health Information
Technology (ONC), shared his perspectives about defining standards and regulations to ensure that health data can be accessible for CBK and subsequent innovation. Dr. Rucker highlighted forces that will ultimately drive the mobilization of CBK, including exponential growth in computing power and data, new devices and consumer adoption, and an "economic revolution" in healthcare. He noted the prevalent national (and congressional) interest in consumer issues, such as drug pricing and surprise billing, and stated that transparent information around price-including analytic methods to compute and communicate price-will be essential to address these issues. Finally, Dr. Rucker noted that the ONC is working to facilitate the exchange of population level data (not just individual data exchange), which can be used by payers to quantify the value of healthcare activities and subsequently incentivize high-quality care.
Dr. Dipak Kalra, President of The European Institute for Innovation through Health Data (i~HD), 13

highlighted several initiatives in
Europe as examples of how CBK can be used to improve research, learning health systems, and population health. Dr. Kalra suggested we learn from data standards communities' efforts to promote the adoption of complex standards. He described the i~HD's "Interoperability Asset Register", which provides potential adopters with various types of resources, including legal (eg, policies and agreement templates), organizational (eg, adoption guidelines and training resources), technical (eg, information models and XML schema), and semantic (eg, clinical models and value sets). 14 Dr. Kalra indicated that these types of resources will enable potential adopters to assess CBK artifacts and support implementation from both technical and organizational contexts. Finally, he challenged CBK developers to view themselves not simply as inventors, but as "founders" of knowledge communities that will continuously monitor and improve CBK.

| Work group action sessions
The speakers described above provided background, vision, and motivation for meeting participants, who were charged to advance the MCBK vision through the four work groups formed during the first MCBK public meeting. Two breakout sessions (90 min each on days 1 and 2) were designated as Work Group Action Sessions. The work groups and their co-chairs, scope, and discussions are summarized below.
The Standards Work Group, led by Bob Greenes and Bruce Bray, is focusing on the identification of informatics standards specifications and metadata needed to support the application and FAIR knowledge capabilities for CBK. The work group is examining several CBK use cases that illustrate various types of CBK artifacts, and plans to identify metadata to characterize CBK and achieve FAIR goals. The metadata will include important dimensions (eg, source, purpose, intended user, domain) related to finding, accessing, interoperating and reusing CBK artifacts. One dimension of the metadata will likely include the data requirements of CBK objects, establishing linkages between standards for representing knowledge objects and clinical data representation standards. To this end, the group is exploring a map of relevant data standards that will expose areas where knowledge developers and implementers need guidance on how to combine various data standards and identify opportunities for harmonization and coordination of standards development efforts. This focus is on identifying and implementing existing standards rather than on developing new ones. Dr. Dymek and Mr. Perry are committed to identifying and including other critical stakeholders, including patient advocacy groups and patients, who represent ultimate end users of CBK, and policy makers, many of whom are aware of the need for oversight or good practices for managing CBK (eg, the proposed "Algorithm Accountability Act". 24 ) As a first activity, the Sustainability work group is using information from four interviews to develop targeted messages and communication strategies for affinity professional societies. Provide infrastructure? Promote policy? Dr. Van Houweling stated that any one of these functions could be useful, but "the community" needs to decide. Whatever that choice, the next step for MCBK will be to define value propositions for various stakeholders, and develop plans to communicate, engage, and work with potential partners and sponsors. Although the MCBK community is just 2 years old, we will inevitably need to consider governance, funding, resources, and future utility and sustainability of MCBK in the bigger world.

| Findings and Impressions (from the authors)
The number and diversity of meeting attendees indicates that MCBK fills an important but broad niche. As planners and participants in the meeting, we witnessed autonomy, vision, and momentum in the MCBK community. Work group chairs and members are enthusiastic to continue their discussions and activities into the next year. There was general support for subsequent public meetings.
Many meeting attendees were new to the MCBK vision and community, and as a result our work groups spent significant time discussing the scope of the work and different perspectives, ideas, and interests of those participating. This is a known and inherent part of the collaborative process. Coalescing multidisciplinary stakeholders toward a common language and goals in a dynamic and complex ecosystem will be an ongoing challenge, but will be worthwhile if we realize a knowledge ecosystem where CBK is applied routinely, safely, and equitably to improve our health and healthcare experience.