Clinical Decision Support

The Harvard Library of Evidence contains graded recommendations and Appropriateness Use Criteria for Clinical Decision Support for use by EMR vendors, IT professionals, and Qualified Provider Led Entities that comply with Protecting Access to Medicare Act of 2014 (42 USC 1395m, “PAMA”), signed into law on April 1, 2014.

The curation and evidence-grading process is based on a framework that has previously been published (Lacson et al, JAMIA, 2016).

Each recommendation is represented as a unit of evidence and users can select content based on the following pre-defined categories:

  •    Diagnosis/Symptom
  •    Source Type (e.g. local best practice, professional society guideline, peer-reviewed article)
  •    Publisher (e.g. ACR)
  •    Choosing Wisely
  •    Professional Society
  •    Imaging Modality
  •    Body Region
  •    Strength of Evidence

Recommendations conform to industry standards and can be downloaded, as follows:

  • Propositional logic as text files (in plain txt or xml)
  • Spreadsheet including terms mapped to standard terminologies (in Excel)
  • Clinical Decision Support (CDS) knowledge artifacts in Fast Healthcare Interoperability Resources (FHIR) specification (in xml)

Recommendations and Appropriateness Use Criteria are translated to single decision rules. Translation of single recommendations, branching recommendations (e.g. flowchart-based), and score-based recommendations to propositional logic is performed manually. These are available as plain text in IF...THEN format.

Terminology mapping to three standard terminologies, The Systematized Nomenclature of Medicine (SNOMED), Radiology Lexicon (RadLex), and the International Classification of Diseases Revision 10 (ICD10) is in publication (Yan et al, AMIA Annual Session Proceedings, 2016).


Decision rules are represented in FHIR specification using the FHIR "decisionsupportrule” resource for sharing knowledge artifacts. This is used in conjunction with the HL7 Clinical Quality Language Specification. Decision rules in FHIR will soon be available for download in xml.

Visit the Center for Evidence-Based Imaging's website to learn more about our collaborative work on CDS.