LOINC Groups vs queries via LOINC/SNOMED CT Expression associations

Home Forums LOINC Groups LOINC Groups vs queries via LOINC/SNOMED CT Expression associations

This topic contains 0 replies, has 1 voice, and was last updated by  raul montilla 7 months ago.

Viewing 1 post (of 1 total)
  • Author
    Posts
  • #21395

    Daniel Vreeman
    Moderator

    We received the following comment from our colleagues at UNMC and I thought it would be useful to comment on here for others:

    I understand the motivation to develop LOINC Groups as I have long been an advocate of the need to query LOINC datasets intelligently.  However I ask you why you are pursuing this strategy when a deployment of a LOINC ontology using the harmonized  SNOMED CT-LOINC Observables concept model supported by description logic will create a far more flexible and computable tool?  Give me the query use cases you want to support and I will show you how the technology preview work can give you answer sets.

    LOINC Groups and our work with to link LOINC with SNOMED CT Expressions are complementary approaches. The LOINC Groups project is unique in several key ways:

    It works across all of LOINC’s content, including domain areas where IHTSDO is unwilling to cooperate on a shared approach (e.g. assessment scales, clinical document titles, radiology procedures, etc).

    We wanted to assign persistent identifiers to the LOINC Groups (e.g. LG* codes) so that they could serve as value set IDs and be plugged into decision support rules, etc.

    We wanted a mechanism to attach annotations and other related information (e.g. molecular weights, conversion formulas) to the group itself and the items within a group to help users decide whether a particular Group was relevant for their use case. What if there is a documented 7% difference in the arterial versus venous measures for a particular analyte? Should you aggregate those results together? Of course, it depends on your purpose.

    Our main focus is on creating clinically relevant and practically useful roll-ups, not purely ontologic hierarchical groupings. These are complementary approaches with slight differences. A hierarchy based purely on ontologic relationships could help group all tests for a particular analyte on various subtypes of blood, but would not account for the fact that you’d always treat tests on cord blood or dried blood spots separately or that the levels of some substances change in various parts of the circulation (e.g. Arterial versus venous). So while the ontology relationships are certainly useful and worth pursuing, we wanted a mechanism that would allow us to capture these other kinds of distinctions too.

Viewing 1 post (of 1 total)

You must be logged in to reply to this topic.