On Sunday 16 May 2004 21:08, Mary Hawking wrote:
> However, my understanding is that whereas Read is hierarchical - and
> SNOWMED isn't.
The version of Read (CTT whatever) involved is version 3.1 (or later if there
is one)
Read 3 is not hierarchical.
Read 3.1 is Read 3, obviously, with all codes and terms from the 5 byte set of
Read 2 and all from the 4 byte set as well, added to it. Most of them have
an indication they are deprecated, and point to a preferred newer code, which
obviously should have the same or an equivalent rubric.
Any hierarchy you like may be impressed on Read 3 - or several hierarchies.
For instance a hierarchy in which routes from
HT :: secondary :: due to drug effect :: contraceptive pill
and
Contraception :: adverse effects :: hypertension
arrive at the same code and rubric, being the same concept, and may be
searched out by the same search. (and the Tb example, yes)
(What the geniuses who were doing the code=selection and rule-writing for the
current variant of our contracts understand of this is wholly unclear to me,
but no doubt they will invent a new set of codes to use for their purposes.)
There is some clever stuff about the distinction between classification
systems and coding systems, but its late, and what we really need is an
ontology, hidden well out of sight.
--
Adrian Midgley (Linux desktop)
GP, Exeter
http://www.defoam.net/
|