Fast Evidence Interoperability Resources (FEvIR) Platform

The FEvIR Platform enables creation and viewing of computable (precise, unambiguous) expression of scientific knowledge in global standard form.

Brian S. Alper, MD, MSPH, FAAFP, FAMIA

CEO, Computable Publishing LLC

Project Lead, EBMonFHIR

Project Lead, COVID-19 Knowledge Accelerator

• We solve the information infrastructure problem needed to solve the problem of Respond: decrease transmission and spread by identifying measures that are effective, equitable, and affordable.
• To identify measures that are effective, equitable, and affordable (and to do this efficiently on a global scale with thousands of dissimilar collaborators) we need consistent agreement (standard) in how measures are defined, how populations are defined, how outcomes are defined, how effectiveness is defined, how equitability is defined, and how affordability is defined.
• We provide tooling (user interfaces, APIs) and platforms for machine-interpretable expression of evidence (the definition of the variables, the statistics related to that variable set, and the certainty of these findings) and related concepts (such as study design, citations, and recommendations).
• A common computable form of scientific knowledge communication enables efficiency of sharing and re-use of knowledge.

The primary users of FEvIR.net are scholars, researchers, and information specialists. Original investigators may use FEvIR.net to report their results in computable form. Evidence evaluators (such as systematic reviewers) may use FEvIR.net to report their critical appraisal of the scientific knowledge, reporting their "certainty" in the knowledge. Guidance developers may use the computable expression of scientific knowledge to more quickly find and more quickly be alerted to the evidence of greatest relevant to their guidance.

Any member of the target audience is welcome to join the COVID-19 Knowledge Accelerator (https://confluence.hl7.org/pag...) for open, free collaborative efforts as we apply these tools to advancing COVID-19 knowledge.  We also set up a low-cost collaborative Apollo Accelerator Membership Program (http://computablepublishing.co...) for organizations that want to contribute more directly to the tools development.

Accelerating communication of scientific knowledge will enable immense "public good" results.

Researchers and consumers of research results (including decision-makers of many types) find communication of, evaluation of, and finding research results intensely difficult and time-consuming.  Our solution will enable computable tooling (including artificial intelligence and machine learning) to improve all of these workflows and overcome domains where our knowledge has not previously been expressed in precise, unambiguous terms that machines can use.  Natural language processing is inadequate becomes the scientific community is not consistently accurate in communicating research results in natural language.

Apollo Accelerator members will pay for influencing feature development on the FEvIR Platform. The users will thus continually direct development to make it more useful which in turn increases the value proposition.  The pace of development will scale with the volume of use.

We are currently setting up the base platform and will make specific measurable indicators with specific projects developed upon the platform.  There will be multiple different projects with different collaborators. Each project will have project-specific metrics. For example we recently submitted a proposal for a grant award to develop a "COVID-19 Trial Results Mapper". In that project, the metrics to demonstrate feasibility are:

1. Two independent researchers use the data entry forms to enter data from two clinical trials, and the data summaries produced are confirmed by a third independent researcher to match the original trial results.

2. Five independent researchers use the data entry forms to enter data from two clinical trials, and they complete the task without requiring technical support or additional instruction.

3. Implementers from at least two independent systems intend to continue using the data entry forms for efforts not supporting by the grant funding.

We are establishing the infrastructure for an immensely scalable method for standardizing electronic communication of science.  The major initial barriers are acceptance of the standard, in terms of agreement and adoption.

For agreement with the standard, we are working through HL7, a standards developing organization, to follow proven methods for global agreement. We have been able to establish an "Accelerator" approach that greatly increases the pace of standard development and initial testing. For $25,000 per year we could formalize the Accelerator program with HL7 for greater integration with HL7 processes and networking.

The biggest barrier to adoption is initial adoption.  Once a few systems use the standard, the efficiency of communication increases so strongly that additional growth will be "viral" as system users directly expand use and community through the communication. The critical need to overcome is an initial community to demonstrate functionality. We could find initial users in many sectors including digital publishers, clinical trial management systems, clinical decision support developers, and research organizations. We are currently considering systematic reviewers and clinical trial registry systems for initial efforts. These groups are selected for convenience and we are open to work with other groups in the initial stages.

Computable Publishing LLC

COVID-19 Knowledge Accelerator (COKA) Initiative

Health Level 7 International (HL7) -- through EBMonFHIR project

One of the Members of The Trinity Challenge encouraged us to apply.   Our solution will provide a platform for interoperability for thousands of applications and platforms. As such many other collaborators can become customers, contributors, or partners. We are also producing such a societal impact that the award itself could help overcome the initial barrier to make it happen.

Doctor Evidence is a The Trinity Challenge Member organisation that is obviously directly interested in a standard structure for reporting evidence and would benefit greatly from the FEvIR platform.

There are likely many other member organisations that would find our efforts to add value.

The simplest model for partnering is for organisations to join the Apollo Accelerator Membership Program (https://computablepublishing.c...) which provides them a low-cost method to support the efforts and license the software we develop. We are open to other types of partnerships as long as they are not restrictive with respect to be supporting and enhancing global communication of science.