Skip Navigation Archive: U.S. Department of Health and Human Services U.S. Department of Health and Human Services
Archive: Agency for Healthcare Research Quality
Archival print banner

This information is for reference purposes only. It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work. Persons with disabilities having difficulty accessing this information should contact us at: Let us know the nature of the problem, the Web address of what you want, and your contact information.

Please go to for current information.

Clinical Decision Support Consortium (Text Version)

Slide presentation from the AHRQ 2008 conference showcasing Agency research and projects.

Slide Presentation from the AHRQ 2008 Annual Conference

On September 8, 2008, Blackford Middleton, M.D., M.P.H., M.Sc., made this presentation at the 2008 Annual Conference. Select to access the PowerPoint® presentation (2.2 MB).

Slide 1

Clinical Decision Support Consortium

Blackford Middleton, M.D., M.P.H., M.Sc.
Clinical Informatics Research & Development
Partners Healthcare
Brigham & Women's Hospital
Harvard Medical School

Slide 2

AHRQ CDS Demonstration Projects


  • To develop, implement, and evaluate projects that advance the understanding of how best to incorporate CDS into health care delivery.

Overall goal

  • Explore how the translation of clinical knowledge into CDS can be routinized in practice and taken to scale in order to improve the quality of healthcare delivery in the U.S.


  • $1.25 million per project per year for two years.

Slide 3

CDS Background

  • Clinical decision support (CDS) has been applied to:
    • increase quality and patient safety.
    • improve adherence to guidelines for prevention and treatment.
    • avoid medication errors.
  • Systematic reviews have shown that CDS can be useful across a variety of clinical purposes and topics.

Slide 4

Barriers to effective CDS

Current adoption of advanced clinical decision support is limited due to a variety of reasons, including:

  • Limited implementation of Electronic Medical Record (EMR), computerized physician order entry (CPOE), personal health record (PHR), etc.
  • Difficulty developing clinical practice guidelines.
  • A lack of standards for knowledge representation.
  • Absence of a central repository for knowledge resources.
  • Poor support for CDS in commercial Electronic Health Records (EHRs).
  • Difficulty tailoring CDS to context of care.
  • Challenges in integrating CDS into the clinical workflow.
  • A limited understanding of organizational and cultural issues relating to clinical decision support.

Slide 5

The CDS Consortium Primary Goal

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.

Slide 6

CDS Consortium: Founding Member Institutions

  • Partners HealthCare.
  • Regenstrief Institute.
  • Veterans Health Administration.
  • Kaiser Permanente Center for Health Research.
  • Oregon Health and Science University.
  • University of Texas.
  • Siemens Medical Solutions.
  • GE Healthcare.
  • MassPro.
  • NextGen.

Slide 7

Six Specific Research Objectives

  • Knowledge management lifecycle.
  • Knowledge specification.
  • Knowledge Portal and Repository.
  • CDS Knowledge Content and Public Web Services.
  • Evaluation process for each CDS assessment and research area.
  • Dissemination process for each assessment and research area.

Slide 8

CDS Consortium Teams chart

The chart outlines the lead and teams for the following:

  1. KM Lifecycle Assessment.
  2. Knowledge Translation and Specification.
  3. KM Portal and Repository.
  4. Vendor Generalization and CCHIT.
  5. CDS Services Development.
  6. CDS Demonstrations.
  7. CDS Dashboards.
  8. CDS Evaluation.
  9. Masspro Dissemination.
  10. Joint Modeling Working Group.

Slide 9

Workflow Diagram

The diagram shows the Workflow process.

Slide 10

Multilayered Knowledge Representation

  • Provides balance between the competing requirements for flexibility in representation for various IT environments,. and the ability to deliver precise, executable knowledge that can be rapidly implemented.
    • Make available a machine executable level knowledge artifact for where it can be used (easy implementation, rapid updates).
    • For others, it may be more appropriate to use an artifact from the Semi-structured Recommendation or Abstract layers, to allow rapid implementation of their own executable knowledge.
  • Provides a path to achieve logical consistency from the narrative guideline to the execution layer.

Slide 11

Multilayered model

The table shows a multilayered model comprised of Machine Execution, Abstract Representation, Semi-structured Recommendation, and Narrative Guideline. Two arrows, one pointing up: Precision and executability; one pointing down: flexibility and adaptability are on either side of the table.

  • Machine Executable layer:
    • Knowledge encoded in a format that can be rapidly integrated into a CDS tool on a specific HIT platform.
    • E.g., rule could be encoded in Arden Syntax.
    • A recommendation could have several different artifacts created in this layer, one for each of the different HIT platforms.
  • Abstract Representation layer:
    • Structures the recommendation for use in particular kinds of CDS tools.
    • Reminder and alert rules.
    • Order sets.
    • A recommendation could have several different artifacts created in this layer, one for each kind of CDS tool.
  • Semi-structured Recommendation layer.
    • Breaks down the text into various slots such as those for applicable clinical scenario, the recommended intervention, and evidence basis for the recommendation.
    • Standard vocabulary codes for data and more precise criteria (pseudocode).
  • Narrative Recommendation layer:
    • Narrative text of the recommendation from the published guideline.

Slide 12

Knowledge Pack

  • For each knowledge representation layer in CDS stack:
    • Data standard (controlled medical terminology, concept definitions, allowable values).
    • Logic specification (statement of rule logic).
    • Functional requirement (specification of IT feature requirements for expression of knowledge—rule, order set, template, etc.).
    • Report specification (description of method for CDS impact measurement and assessment).

Slide 13

Complete CDS Knowledge Specification

The empty table shows the Multilayered Knowledge Model layers(Narrative, Semi-structured, Abstract, and Machine interpretable) represented in the left column with the Knowledge Pack layers (Data, Logic, Function, and Measure) represented across the top row.

  • A complete functional specification to accommodate and facilitate a variety of implementation methods in HIT.

Slide 14

Complete CDS Knowledge Specification

The previous table filled in:

  • Narrative.
    • Data-if the patient's creatinine.
    • Logic-is elevated then avoid metformin.
    • Function-ability to show an alert (on screen or paper).
    • Measure-percentage of metformin pts with a high Cr.
  • Semi-structured.
    • Data-lab value: creatinine.
    • Logic-clinical scenario: elevated Dr.; Action: avoid metformin.
    • Function-lab results, medication list (database).
    • Measure-num: all metformin pts; Denom: high Cr & metformin.
  • Abstract.
    • Data-LOIN C 2159-2.
    • Logic-if cr >1.2 mg/dL → tell user "d/c metformin."
    • Function-CIS with rule evaluation capablility, alerting function.
    • Measure-num: NumSet = {med=metformin} DenomSet={cr > 1.2}.
  • Machine interpretable.
    • Data-select * from labs where ID = 2159-2.
    • Logic-if (cr >1.2)?print ("d/c metformin").
    • Function-CPOE with lab, meds and alerting capability.
    • Measure-select count (*) where.

Slide 15

Knowledge Artifacts by Layer

  • Published Guideline.
  • Semi-structured recommendation.
    • Abstract Rule.
      • Executable Rules.
    • Abstract Order Set.
      • Order Sets in CPOE system.

Slide 16

Accomplishments to Date (start 3/08)

  • KM Lifecycle Assessment Team.
    • Completed Knowledge Management and CDS Survey and sent it out to the Consortium sites. PHS and Regenstrief have returned the survey.
    • PHS Site Visit, June 16-20. Interviewed and shadowed Partners physicians about their knowledge management and CDS practices.
    • Site visits to Regenstrief and VA scheduled and shepherds identified.
  • Knowledge Translation and Specification Team.
    • Completed semi structured representation and presented work to AHRQ and TEP on July 11, 2008.
  • Draft clinical action model developed.
  • KM Portal.
    • Delivered eRoom as a collaborative environment for CDSC activities and finalized KM Portal design hardware.
  • Vendor Generalization and CCHIT Team.
    • Completed capability reviews of nine EHR systems through customer interviews to assess their decision support features.
  • CDS Services Development.
    • Completed literature review on current service-oriented architectures for clinical decision support.
    • Beginning service development.
  • Joint Information Modeling Working Group.
    • Patient data model and terminologies selected.
    • Developing conceptual model.
    • Developing localization model.

Slide 17

Timeline Overview

Year I

  • Knowledge Management Lifecycle Assessment.
  • Knowledge Translation and Specification.

Year I-Year II

  • Knowledge Portal & Repository.
  • Evaluation.
  • Dissemination.

Mid Year I-Mid Year II

  • CDS Web Services Development.
  • Vendor Recommendation/CCHIT.

Mid Year II

  • Demo Phase 1: LMR.

Slide 18


The cartoon shows a sitting man looking at a line graph on a wall with an X showing, "you are here."

  • Thank you!
Current as of February 2009
Internet Citation: Clinical Decision Support Consortium (Text Version). February 2009. Agency for Healthcare Research and Quality, Rockville, MD.


The information on this page is archived and provided for reference purposes only.


AHRQ Advancing Excellence in Health Care