Allergy & Intolerance substance September 2017 ballot outline

Back to Allergy & Intolerance Drug Sub-project

• Statement of problem: Substance codes too numerous to manage effectivley; specifically, C-CDA
  • UI design
  • message validation
  • correct level of specificity
  • errors based on

• Goal: short list, heuristic
  • easy to pick 99% of time
  • easy to validate
  • also
    • correct level of specificity based on immunological understanding of cross-reactivity
    • omit junk (fish, seafood)

• Considerations
  • not pharmacovigilance. compare FHIR properties.
  • length - ask about useful length
  • system tbd. so far realm-specific.

• collection
  • describe data (period, source)
  • string not code. prior to map. therefore some junk.
  • duplication
  • weighting
  • distribution
  • mapping threshold. junk, significance.

• issues found
  • boundaries: caffeine, eggs,
  • herbals: no system. if rxnorm, else sct
  • biologics - blood religious. vaccine: cvx too specific (russ . . .)
  • salts no
  • classes: NDF criteria not cross-reactive; use SCT. check Goss.
  • route: salicylates, iodine,
  • cross-reactive:
    • penicillin redefine
  • clinical: iodine, contrast. Keep iodine topical, iodine contrast(?), add high-osmolality
  • metal, influenza, kinds of tape

• QA process
  • heuristic
  • multiples
  • ballot

• findings
  • list
  • dispersion by institution
  • top X% in Y#

• Guidance for implementers
  • Nugatory values with high counts not included (e.g., seafood)
  • Inflated values with high counts included with caveats (e.g., iodine, iodinated media)
  • Application of true values (eggs & vaccine, erithromycin & azithromycin)

• Questions for readers
  • what's the right length
  • one list or many?
  • what do we do about caffeine (RxNorm) vs caffeine (SCT)?
    • one list: which code
    • many lists: how to implement
  • feasibility & usefulness of criticalities