|Subject:||Re: [Gnumed-devel] basic test types - Sample type prefixes|
|Date:||Fri, 08 Oct 2004 19:56:53 +1000|
|User-agent:||Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.7) Gecko/20040616|
Complicating matters - but at some point something we may want to factor into design - is that the labs often use a different set of codes by which to ORDER tests as opposed to codes by which to REPORT results. Among other reasons, a single order code can designate multiple tests whereas that single code would be a clumsy way by which to manipulate/search results. Also there are some tests which although you may order a single test (for example serum iron saturation), I think the lab must do a pair of tests (transferrin and iron) and even if that is wrong there are other situations where they must do a "ghost test" (my term) in order to generate a result.
so what are you trying to say ? It sounds like "this relationship is 1-many , 1 order code to many results code, however only one result code relates to an order code. (i.e. there is a functional dependency between result to order code), so I need 2 tables , not 1 , for test_type. (test_type_order, test_type_result). Are you wanting to put version fields in the both tables ( to accomodate for lab fashion trends)? What about just posting a change as tablename1(attributes) -1- relationship-name -*- tablename2(attributes). Please include in gnumed schema, sir.
Or maybe dig in CDA and see how they do it , and then rename it so it belongs to gnumed. (i.e. re-invent the wheel, that's what a path friend said I was doing when I started my interest in programming, joining the profession that re-invents the wheel for a living , instead of getting a life ).
|[Prev in Thread]||Current Thread||[Next in Thread]|