SNOMED Documentation Search


 Other Documents

Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

Moving from a legacy coding scheme to

Glossary ref
SNOMED CT
requires attention to be paid to continued accessibility and use of data encoded using the legacy scheme.

The following general approaches may be applied or adapted taking accounts of the capability of the

Glossary ref
SNOMED CT enabled application
and the value and relevance of existing data.

  • Mapping or converting the data:
    • This requires each instance of a code in the existing data to be mapped to an appropriate
      Glossary ref
      SNOMED CT expression
      . This
      Glossary ref
      expression
      is then associated with the existing coded record entry as part of a record entry that conforms to the information model used in the
      Glossary ref
      SNOMED CT enabled application
      .
  • Linking or integrating existing data:
    Anchor
    _926f3c70-3e91-4a8f-8d79-08cde2ca13f8__a
    _926f3c70-3e91-4a8f-8d79-08cde2ca13f8__a
    • The existing data is retained in its native form or in an intermediate form. However, the
      Glossary ref
      SNOMED CT enabled application
      is designed (or adapted) to access this existing information as if it had been converted. To deliver this functionality, queries and/or data are in effect mapped at the time of retrieval, rather than at the time of upgrading the system. This can be achieved in different ways some of which involve direct use of mapping tables and others in which, while the existing data is unchanged, an index derived from a map to
      Glossary ref
      SNOMED CT
      is generated to optimize subsequent access.
  • Archive or retain old data in its original form, and where it is necessary to retrieve historical information, use
    Glossary ref
    components
    from the legacy system to do this:
    • This approach completely separates the new
      Glossary ref
      SNOMED CT
      from the legacy data and is unlikely to be acceptable in clinical practice . However, it may be appropriate for some data warehousing applications where the wholesale conversion of data is considered too onerous.