May 20

May 20 QRPH Face-to-Face Meeting Minutes

Attendees:

• Floyd Eisenberg (Siemens) Co-Chair
• Jason Coquitt (Greenway Medical Systems) Co-Chair
• Landen Bain (CDISC)
• Ana Estelrich (GIP-DMP) - Planning Co-Chair
• Joann Larson (Kaiser)
• Judy Logan (HL7, OHSU)
• Joan McMillen (RSNA)
• Harry Solomon (GE)
• Gary Walker
• Daemon Whittenberg (Greenway Medical Systems)

1. Measure Definitions with Value Sets
   1. Overview of knowledge to guideline to clinical decision support to measure model Diagram courtesy Sheila Teasdale, AMA
   2. Collaborative model may not be sufficient for the expression of quality measures. It does not manage continuous variables (mean, median). Counts for population or denominator are managed, but not the continuous variable.
      1. Context must be identified to disambiguate the source:
      2. The section of the CCD or summary to represent the information will attribute different meaning. E.g., Disease may be negated in present history, represent risk in the family history and represent past occurrence in the past medical history.
      3. CCD doesn’t manage procedure reporting. CDA constrained may help but there is no clear gastroenterology procedure constraint on CDA.
      4. Element to identify “withdrawal” time for colonoscopy is not clear. There is narrative that the “mean withdrawal time” is required for reporting. The time (in narrative) is from the time at which the cecum is reached to the time of withdrawal (expected > 6 minutes). How such times are expected to be documented is not clearly specified in the measure.
      5. The measure is only valid for “normal” results for patients with intact colon.
         1. What is “normal” – no polypectomies, no biopsy are likely requirements – Electronic definition will require assumptions
         2. What is “intact colon” – no prior resections, etc. – Electronic definition will require assumptions
      6. White paper will include recommendations for enhancement of such measures:
         1. Review of measures for detailed specification to manage issues identified
         2. Request for new SNOMED codes (perhaps)
         3. Request from appropriate GI expert societies for new structured document or other standard procedure report for Gastroenterology procedures and/or specifically Colonoscopy. Such requests could be made to HL7, DICOM, IHE appropriate domain.
2. Value set Issues:
   1. Registry – Source of creating value sets. The source will send value sets to a repository so that all metadata, oids, identifiers, versions are available for queries to find the appropriate value set. For Joint Commission measures that update every six months, there is a need to identify earlier versions of value sets for analysis of performance during previous time frames.
   2. Effective Time – Important to know a range when a value set is initially valid (effective) and when it is no longer valid (expired) and needs to be replaced with respect to versioning. Example with Joint Commission measures there is an effective date for measure modifications (hence value sets) and for expiration and updates (versions) to measures and value sets.
   3. Language – A value set in a country with two languages (e.g., Canada) requires representation. Language is a preferred string. The value set is represented by the concepts. There may be different preferred strings within the value set, each string representing a different language.
   4. Obsolete Terms – The value set may have members that have a status of ‘deprecated’ or ‘obsolete’. This issue could be managed with versioning. For example, a medication is no longer valid (removed from the market), or a laboratory test is no longer available.
   5. Versioning -
   6. Validation / Security of Value Sets – Potential for incorrect measurements
   7. Creation of a Value Set
   8. Mapping to local terms is not in scope for value sets specifically but an implementation issue for incorporation of quality measures within EMR Systems.
   9. A hierarchically modeled value set would allow use of different levels depending on the implementation site. Relationships to underlying vocabularies remain within those vocabularies.
3. Clinical Research - Working session for mapping CDASH and CCD