Goals and Objectives

Goal

The goal of the CDS Consortium is to assess, define, demonstrate, and evaluate best practices for knowledge management and clinical decision support in healthcare information technology at scale – across multiple ambulatory care settings and EHR technology platforms.

Six specific research objectives:

All 6 specific research objectives will be assessed, defined, demonstrated, and evaluated in two practical areas of implementation for clinical decision support services. For each research objective, the analysis will focus on issues pertaining to both a) USPSTF Preventive Care guidelines, and b) best practices for two chronic care conditions: adult-onset diabetes, and coronary artery disease.

1. Knowledge Management Lifecycle

   This research will focus on defining best practices for the entire knowledge management lifecycle from clinical practice guidelines to actionable clinical decision support. Current knowledge management lifecycle practices will be surveyed from clinical sites in the Clinical Decision Support Consortium to assess the array of activities across academic and vendor best practices. This assessment will focus on the two demonstration areas: USPSTF Guidelines for preventive care, and the chronic care conditions diabetes and CAD.

2. Knowledge Specification

   This research will focus on defining best practices for knowledge representation, data representation, and specification of knowledge content formats for both human readable expression of content, and expression of content for web services implementation. An assessment and comparison of best practices among members of the Clinical Decision Support Consortium will be performed as part of this research. The Consortium will then define an idealized knowledge representation format, and preferred data representation standards for medications, healthcare problems and conditions, allergies, advanced directives, and laboratory data. The research and knowledge specification will also include specification of a general-purpose architecture for public web services to be used both by members of the Clinical Decision Support Consortium for the purposes of this research and demonstration projects, as well as potentially other vendors and organizations employing healthcare information technology.
   

3. Knowledge Portal and Repository

   To achieve adoption of clinical decision support services at scale, a knowledge portal and repository will be created by the Clinical Decision Support consortium. The knowledge portal and repository will be used by members of the Consortium to perform collaborative knowledge engineering tasks as part of the knowledge management lifecycle to create consensus driven knowledge artifacts for use in the clinical decision support demonstration projects. The Knowledge portal and repository will serve to post both human readable knowledge artifacts for decision support, as well as knowledge specifications in web services form. A key feature of the knowledge portal and repository is that it will serve to aggregate knowledge content for the demonstration project areas from diverse members of the Consortium. Clinical knowledge content for decision support will be gathered from Partners Healthcare, Regenstrief Institute, Kaiser Permanente, and the Veterans Health Administration through survey assessment of their clinical decision support content in the demonstration project domains and incorporation into the knowledge portal and repository. The knowledge portal and repository will use commercial software, the Documentum product, from EMC, Inc., and build upon the Knowledge Management Portal developed already at Partners Healthcare (a new instance will be created for the purposes of this research).

4. CDS Knowledge Content and Public Web Services

   To broadly implement clinical decision support services at scale the knowledge portal and repository will make available to clinical members of the CDS Consortium human readable knowledge artifacts for the demonstration projects. This will facilitate rapid adoption of clinical decision support by decreasing the burden of the end user adopting electronic health records and implementing well-specified knowledge content for healthcare maintenance and preventive care services, as well as the chronic disease conditions selected for the demonstration projects. In addition, CDS demonstrations will include creation of publicly available web services in the domain areas of the demonstration projects for consumption by both remote academic electronic health records at the Regenstrief Institute, and the Veterans Health Administration, as well as by the commercial electronic health record product available from Siemens healthcare.
   To demonstrate and assess the viability of the best practices as defined in specific research objectives 1 -- 4 above two demonstration projects will be defined and deployed across two or more electronic health record systems. CDS demonstration projects will be performed in two areas:

   1. Development of comprehensive CDS services for healthcare maintenance and preventive care screening based upon United States Preventive Services Task Force guidelines for healthcare maintenance and preventive care screening. Based on the best practices defined above, a comprehensive set of knowledge artifacts will be created in human readable form for presentation in the knowledge repository and portal, incorporation in the target EHR applications (LMR, Regenstrief Institute, VHA, Siemens LCR) in either a manual process, or via web services.
   2. Similarly, a second demonstration project will focus on using the best practices and knowledge artifacts as described above in the area of chronic disease clinical decision support in the target EHR applications described above.

   For each demonstration project, a critical component will include development, implementation, and evaluation of a clinical decision support dashboard. The CDS dashboard will inform the end user as to his or her compliance with CDS recommendations produced in the demonstration projects.

5. Evaluation

   A common theme underlying all of the above assessments, best practice definition, and demonstration projects will be an evaluation process for each project. Evaluation of the knowledge management lifecycle and the knowledge specification work will occur through consensus development and peer review. Evaluation of the knowledge portal and repository will occur through user assessment of satisfaction, performance, and efficiency of collaborative knowledge engineering. The CDS services demonstration projects in healthcare maintenance and preventive care screening, and chronic disease conditions, will include both basic evaluation components for proof of concept assessment, usability assessment, pilot implementation, and user satisfaction assessment. Hopefully, follow-up research projects will allow more formal assessment (beyond the scope of this contract’s provisions) in the form of prospective randomized controlled trials. The CDS services demonstration projects occur across at least two and potentially more electronic healthcare record products, including the LMR at Partners Healthcare, the Regenstrief Institute electronic health record, the VHA Vista EMR, and Siemens LCR - Lifetime Clinical Record.

   In addition to the core CDS technology evaluations described above a second thread of evaluation will proceed to assess the requirements for generalization of the best practices defined in this research by members of the CDS Consortium for other vendors of healthcare information technology. This will occur through survey assessment of essential EHR requirements as defined by this research for improved adoption of CDS knowledge content in human readable form, or as publicly available web services.

   A third critical thread for the evaluation objectives in this research will be to identify best practices for deploying clinical decision support services in the context of electronic health records in ambulatory care practices. This will occur through survey assessment of organizational barriers and enablers, and assessment of best practices the CDS Consortium membership, as well as qualitative assessment of the implementation and deployment of the CDS demonstration projects.

6. Dissemination

   After completing the four research objectives mentioned above, we will disseminate our findings.  For this, we will conduct analysis to produce academic publications and presentations, as well as the appropriate reports to the funding agency AHRQ.

   A second method of dissemination for the findings of this research will be to make insights and knowledge artifacts available to the National Resource Center for Healthcare Information Technology, sponsored by the AHRQ. The NRC will create abstracts of best practices for the knowledge management lifecycle, knowledge specification, CDS implementation and deployment for CDS services, available on its website.