Summary of third annual MCBK public meeting: Mobilizing computable biomedical knowledge—Accelerating the second knowledge revolution

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 was formed in 2016 to address these needs. This report summarizes the main outputs of the third annual MCBK public meeting, which was held virtually from June 30 to July 1, 2020 and brought together over 200 participants from various domains to frame and address important dimensions for mobilizing CBK.


| BACKGROUND
Despite the growing pace of biomedical knowledge, 1 the United States continues to see increasing health disparities and decreasing life expectancies for some populations. 2 The number of potential treatments and evidence base is steadily increasing but it is difficult to disseminate and implement these into practice, due to a number of challenges including volume, relevance to particular patients, and need to adapt to particular workflows and EHR technologies used where health-related decisions are made. To be widely disseminated and actionable, knowledge needs to be distributed in usable and implementable formats-ie, computable biomedical knowledge (CBK). CBK, such as predictive models, rules, alerts, clinical pathways, or data visualizations, is necessary for the interventional approach of a learning health system. The mobilization of CBK can result in rapid mass access to computable knowledge with the potential to improve the health of individuals and populations on a large scale. 3 We believe that a new, coordinated ecosystem is needed to revolutionize how knowledge and evidence can be distributed to support decision-making and action and thereby benefit human health.
While work has proceeded in this space for many years, the movement to mobilize computable biomedical knowledge (MCBK) was conceived 3 years ago by a number of thought leaders. The MCBK movement aims to achieve better health in diverse settings by widely sharing knowledge in a computable format. 4,5 In this report, we present a summary of the third annual MCBK public meeting held virtually from June 30 to July 1, 2020.

| MEETING AND PARTICIPANT INFORMATION
Due to the COVID-19 pandemic, the meeting was held as an online interactive conference, replacing the live meeting originally planned to take place at the National Institutes of Health in Bethesda, Maryland.  Also, their many activities/projects that support the data, tools, methods, and policies around AI and ML to ready our nation for that change. Finally, she shared plans of the NLM to support data access and knowledge creation activities through a number of Data Design Centers, which would ensure data quality and access, transparency about bias, and development of tools to enable/ empower more consumers to generate knowledge to address persistent and emergent health problems.  Dr Herbert Sauro described the way biomedical models are currently published (or not!), emphasizing that the future should be long-term model repositories and technologies to manage published models-with adequate information and metadata to be reproducible. He likened the current human readable journal article to an "advertisement" for a study or model description but noted future implementers and health information consumers need more resources to evaluate or apply these models. Dr Sauro called PubMed a host that could lead the move to the digital preservation of models.

| WORKGROUP ACTION SESSIONS AND ACTIVITY
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. One breakout session (3 hours on Day 2) was designated as a Work Group Action Session. The work groups and their co-chairs, scope, and discussions are summarized below.
The Standards Work Group SWG), led by Drs. Robert Greenes and Bruce Bray, is focused on identifying existing and emerging standards that will facilitate widespread use of knowledge by enabling FAIR 8 and Trust capabilities. 7 During the breakouts at the meeting, the SWG discussed efforts to define a set of nonoverlapping categories of metadata that will enable the FAIR+T capabilities. Prior to the SWG session, a team had been involved (ongoing since then) in defining these categories and using exemplar artifacts of different types to help elucidate them. It should be noted that the categories, which included Type, Biomedical Domain (now called Domain), Purpose, and Coverage (now called Evidential Basis), each have multiple dimensions of metadata, so a task is that of clarifying what those dimensions are, which is best done by examining artifacts. Also note that 7 other categories, for a total of 11 were briefly presented. After an introduction to the challenge, participants were asked to join separate sub-breakouts, each examining the four categories, and to report back at the