Details

    • Change Description:
      Hide
      The following changes are made:
      - the SINGLE_EVENT class is removed, and the remaining class structure
        collapsed back to HISTORY and EVENT.
      - EVENT.offset is replaced by EVENT.time:DV_DATE_TIME, which is the
        absolute time of the event; in the case where width is non-zero,
        it is the trailing time.
      - add functions start_time and end_time which return the start and
        end times of the event - same in the point case, different in the
        non-point case.
      - add state: ITEM_STRUCTURE to EVENT.
      Show
      The following changes are made: - the SINGLE_EVENT class is removed, and the remaining class structure   collapsed back to HISTORY and EVENT. - EVENT.offset is replaced by EVENT.time:DV_DATE_TIME, which is the   absolute time of the event; in the case where width is non-zero,   it is the trailing time. - add functions start_time and end_time which return the start and   end times of the event - same in the point case, different in the   non-point case. - add state: ITEM_STRUCTURE to EVENT.
    • Approved By:
      ARB

      Description

      There are various proposals to change HISTORY<T>, as follows:
      - instead of using origin & offsets, use absolute times in each EVENT
      - remove SINGLE_EVENT and simplify class structure back to just
        HISTORY and EVENT
      - make it optional (AG) because many instances of OBSERVATION will
        not specifically record their own observation time - they
        'inherit' the time of the clinical session. (TB: I agree with
        the point - in today's GP consults - since there is almost no
        automated recording of anything, but disagree with the solution.
        Recording of observation time is essential, and even if the
        absolute times of multiple observations in a GP consult are
        uninteresting, the order may be significant.

      A further requirement which has come to light during archetype
      and system-building in Australia is that state:ITEM_STRUCTURE
      should be available on EVENT, since it is usual to record
      state at the time of the data event, rather than as a separate
      time series on OBSERVATION, which it is currently. The latter is
      not incorrect and should be retained; but the addition of EVENT.state
      will handle the more common clinical cases.

        Attachments

          Activity

            People

            • Assignee:
              OLDthomasbeale JeffJ
              Reporter:
              OLDsamheard (Sam Heard) inactive (Inactive)
            • Votes:
              0 Vote for this issue
              Watchers:
              0 Start watching this issue

              Dates

              • Created:
                Updated:
                Resolved: