HCLS/ClinicalObservationsInteroperability/TermMap

From W3C Wiki
Participants
Josh Mandel
Gökçe Banu Laleci Erturkmen
Mustafa Yuksel
Emory Fry
Connor Dowling
Eric Prud'hommeaux, W3C
Abstract

For many clinical research or clinical informatics projects, the most expensive opperation is mapping values in one terminology to another. This work is frequently seen as project-specific but has extensive overlap with other similar projects. This task is to create an exchange of project-specific mappings and the parameters that qualify how others may use them.


Strategy

  • Initially scope to one domain: Ontology of diseases.
  • Find the commonalities between two existing systems (SALUS and TAPS)
  • Derive minimal intersecting requirements (and expressivity)
  • Parameterize factors which interfere with identical mappings (e.g. for X disease,

Outreach

Folks we need to engage in this work:

  • FDA working on Therapeutic Areas for clinical trial metadata interop.
  • Pharmas largely unable to find and access similar trial data.

See Also

Workspace

project clinical pharmacological ...
SALUS ICD10GM, ICD9CM, SNOMED CT (Clinical Findings), MedDRA ATC We have partial mappings between ICD9CM to SNOMED CT (based on the mappings provided by OMOP Project). We can share this. We have partial mappings between Snomed CT to MedDRA based on OntoADR Ontology, we are still working on the mappings between ICD10GM to Snomed CT. Due to licensing issues, this cannot be shared directly.
TAPS ICD9 RxNorm Emory and Connor to fill in

Commonalities

Both are using skos:narrower to capture hierarchy placement and skos:exactMatch for identical concepts.

Context Mismatch

Term1 predicate Term2 context ...
ICD9:123.4 exactMatch SNOMED:1234567 context1
SNOMED:1234567 exactMatch Medra:98765432 context2
ICD9:123.4 exactMatch Medra:98765432 too inclusive, e.g. heart attack coded in ICD9 would be athlete's foot in Medra

Sources

name domain from to description
OMOP drug outcomes ICD9 SNOMED project cataloging the outcomes of medications