Re: Metathesaurus issues. was: Re: does RDF require understandin

> But this problem is a huge one for the healthcare field. Things like "master
> patient index" are good examples of real world thesauri which map so-called
> unique identifiers (e.g. patient id) from one part of an institution to
> another. Having taken part in the merging of several healthcare
> organizations I can assure you that its all alot of fun (and good healthy
> human readable prefixes to UIDs such as are found with URLs are a really
> huge help in these cases).

One thing that looks sound (but I've no specific experience with it) is
*corroboration* as described in HL7's "Context Management (CCOW) Specification
Technology- and Subject-Independent Component 6 Architecture Version CM-1.3 7"
at http://www.hl7.org/Memonly/ballots/ccow/v13_file/hl7_ccow_arch_cm_1_3.pdf

   """Context data items are grouped by subject, wherein each subject contains
      a semantically related set of data items. There are two classes of subjects:

      - An identity subject identifies a real-world entity (e.g., a specific
        patient) or concept (e.g., a specific encounter).
      - An annotation subject contains data that describes or is otherwise
        pertinent to the entity or concept identified in an identity subject.

      The context data items from which a subject may be comprised may
      represent the following categories of information:

      - data that identifies a real-world entity or concept (such as a
        specific patient or a specific encounter),
      - data that can be used to *corroborate* the identity data,
      - data that annotates the identity data with additional information
        from an authentic source that describes the identified entity or concept.

      Identity information is required in order to establish a common context
      between applications that involves a real-world entity or concept.
      *Corroborating* data can be used by applications and/or users as a basis
      for checking further that the identified entity or concept is what was
      expected. Annotation data can be used by applications as a means for
      interchanging data pertaining to the identified entity or concept, wherein
      this information is provided by an authentic data source designated by a
      site.

      For example, a patient's name can rarely be used to uniquely identify
      a patient. Typically, a medical record number or similar identifier that is
      generally unique over some population of patients for one or more clinical
      systems is used. However, these identifiers are rarely meaningful to the
      user. *Corroborating* data might be comprised of the patient's name, sex,
      and data of birth. This data provides applications and/or the user with an
      additional means to check that the identified patient is the intended patient.
      Annotating data might be comprised of the patient's marital status, home
      telephone number, and nationality. This data provides applications with a
      means to share data that describes a patient in addition to identifying the
      patient.

      The clinical context is considered to have changed in a meaningful manner
      when identifier data is set. Applications are notified of changes to the
      context when identifier data, and possibly *corroboration* and/or annotation
      data, are set. Changes to *corroboration* data and/or annotation data that
      are not accompanied by associated changes to identifier data are not
      meaningful and are rejected.
   """

--
Jos De Roo, AGFA http://www.agfa.com/w3c/jdroo/

Received on Tuesday, 13 February 2001 11:01:09 UTC