2023-03-17 Immunization Innovations Group Meeting notes

FHIR-40267 - Getting issue details... STATUS Defer to when Clément is on the call. 

FHIR is going through text and there are small grammar corrections. Good to go for R5. 

• Need to update the workflow description for ImmunizationRecommendation (maybe in Immunization too)
  • Talked to Brynn Rhode at the last WGM
  • He liked the diagram but thought the narrative needs to be updated. 
  • Nathan will work on language and diagram for the group to review, Nathan completed a draft: Immunization vs Medication Administration Resource
  • 20230117 Q4 PHWG WGM Minutes
• We may not be able to get the changes in R5
  • There is a freeze deadline Friday Feb 24 (technical changes) 
  • And one later in March (all content)
  • If we agree on content, we may be able to get it in to R5, but it may be tight
  • Still worth doing sooner rather than later even if we miss R5
• Everyone should look at the page and add comments
• We need a better term for "Immunization System" in the blue box
  • This is really all "non-clinical space" - could be public health, patient facing, community spaces
• We may want to remove the Medication Dispense and Medication Administration boxes to avoid being too specific

Nathan will create a couple of derivations of this and share in two weeks. 

Do we have a priority for this? 

Do we have a market feel what percentage of solutions are being used? What about school rules? Not really decision support, but more logic that comes after CDS. There is a need for robust clinical decision support for special conditions, as an API. 

What about queries back from DS to the clinical application to get ancillary information. 

Feels like two distinctly different responsibilities. Service that can take all things, truly could make a decision. The other how do you actually get all that information?  

Ideal would be to have a CDS for immunizations service for the US like Smarty. First we have to create the standard, that is consistent with that ideal. High risk conditions needs to be solved. The concern is missing information. 

The possibility of using the flags, either send the medications/details or just flags. Need to be able to identify who gave the information. From CDS engine would rather have the decision coming inbound rather than make that decision. The clinician who gets the result back they have to think why does the CDS engine think the patient is immunocompromised? Ideally we have a line drawn between immunization and other clinical decision. The burden for the definition of degree of immunocompromised rests with the providers. Used to give flu shots for "high risk" patients, but each article on these had different definitions. Can't even compare research results because you are talking about different populations. 

On the immunization side we are rules based. Don't see AI coming into this space. 

American Academy of Pediatrics efforts to improve computability and remove ambiguity from policy statements and recommendations:

Next step:

• Put together some diagrams and information about Questionnaire. 

Questionnaire pushes the complexity out towards the clinician and clinician systems.