Skip to end of metadata
Go to start of metadata



  • Craig Newman

  • Heather Patrick
  • Nathan Bunker
  • Daryl Chertcoff
  • Eric Larson

Discussion items

CDS Hooks & FHIR Connectathon

What should we be doing at FHIR Connecatathon this year? 

  • Ken still thinks that CDS Hooks is appropriate
    • We still have concerns that this isn't quite aligned with our use cases
  • The work on clinical practice guidelines might be more aligned with what we are trying to do
    • This is still in ballot recon and hasn't been published yet 
  • Do we think we should be testing either of these at a future connectathon?
    • We need to talk to Bryn about these standards but we shouldn't delay publication to do so
Publication Plans
  • All ballot recon is complete
  • Nathan is still implementing changes to the spec
    • May make some progress next week
  • When we're ready, we'll need a Publication Request
    • PH WG - aim to get an updated version ready for review at the end of February
    • FMG
Value sets
  • List of Value Sets which still need work:
      • ImmunizationRecommendation.recommendation.forecastReason and forecastStatus
        • Forecast Status and Reason Value Sets 
        • Daryl has a more comprehensive (but likely still not complete) list of reasons
        • These are partially overlapping with the CDSi forecast statuses
          • Not Complete
          • Complete
          • Immune
          • Contraindicated
          • Aged Out
        • The base resource example value set include:
          • Due
          • Overdue
          • Immune
          • Contraindicated
          • Complete
        • Nathan will get the list from the TCH forecaster (Now Lone Star Forecaster)
          • Forecast status from Lone Star - compared with Base Resource
            • due later (deprecated) - 
            • due (deprecated) - Due
            • overdue (deprecated) - Overdue
            • not complete - Not Complete
            • complete - Complete
            • finished - Aged Out
            • contraindicated - Contraindicated
            • finished for season (too late to recommend flu) -
            • complete for season (protected for flu this season) - 
            • assumed complete or immune (adult assumed to have completed MMR series in childhood) -
          • Dose status from Lone Star
            • valid
            • invalid
            • missed
          • Dose reason - not coded but there are general templates for human text:
            • (vaccine name) given (date) is invalid because:
              • before valid date
              • (black out reason)
              • invalid when given same day as (vaccine name)
        • We should reach out to Envision, Cerner, Epic, etc who have CDS engines to see what they have for statuses
          • Craig will see if Eric and reach out
        • We will have to figure out if we create a superset of values or try to focus in on a set of core values
          • We will also have to draw a line between "status" and "status reason"
        • Decision on value sets
          • Forecast Statuses (Preferred Binding):
            Not Complete
            Not Recommended
            Aged Out
            Conditional - not all systems have implement the risk based recommendations and a given recommendation may need provider input or requiring Shared Decision Making (will come back to this next week to finalize if this is one concept or two)

            Forecast Reasons (Example Binding):
            Patient Series is Complete
            Not Recommended Due to Immunization History
            Patient Has Exceeded Maximum Age
            Past Seasonal Recommendation End Date
            Complete for the Season

            Note that there isn't general concensus about the set of codes (thus it's still just an example binding)
            Call out the ability to use the text element of CodeableConcept if there isn't a coded reason available

Action items