gnumed-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gnumed-devel] Data pack for ICD-9 :-)


From: Karsten Hilbert
Subject: Re: [Gnumed-devel] Data pack for ICD-9 :-)
Date: Tue, 27 Sep 2011 23:00:29 +0200

> > I agree ("any one language - same code/term not multiple times in the
> > reference table") but I fail the see the connection with the above
> > statement (we want to search on codes) ?
> 
> I am looking for the method by which to manage (prevent) 
> 
> >     ("any one language - same code/term not multiple times in the
> >     reference table")

Have you tried provoking that state ?

Maybe I need to re-check but that's what the constraints
on ref.coding_system_root are for.

> at the same time as deciding on why
> (and how to best) host multiple language versions of the same
> code in coding.

> Maybe there should be *no* term in
> 
>       ref.icd9
> 
> but only the fk and pk and code, instead normalizing each per-language
> term into
> 
>       ref.Icd9_terms
> 
>               fk_icd9_code
>               language
>               term
> 
> and filtering of the terms that are to appear in the phrasewheel can be on
> language-per-locale.

But then the coded terms can *already* be filtered on data_source.lang
(if that makes sense for the data source and thus .lang is non-NULL).

> Consider – as a use-case for terms in more than one language – a
> multi- or mixed-lingual free-clinic volunteer praxis (say english, french,
> chinese simplified)
> 
> My command of these languages is not equal. When my francophone colleague
> has been managing her familiar-to-her patient with hypertension, then I
> expect that when my colleague performed code attachments, she would rather
> that her search on codes for
> 
>       hyperten
> 
> would yield
> 
>       401     Hypertension artérielle essentielle
>       402 Cardiopathie hypertensive
>       403 Néphropathie hypertensive
>       404 Cardiopathie hypertensive et la maladie rénale
>       405 L'hypertension secondaire
> 
> not
> 
>       401 Essential hypertension
>       401     Hypertension artérielle essentielle
>       402 Hypertensive heart disease
>       402 Cardiopathie hypertensive
>       402 高血圧性心疾患
>       403 Hypertensive renal disease
>       403 Néphropathie hypertensive
>       403 高血圧性腎疾患
>       404 Hypertensive heart and renal disease
>       404 Cardiopathie hypertensive et la maladie rénale
>       404 高血圧、心臓と腎臓疾患
>       405 Secondary hypertension
>       405 L'hypertension secondaire
>       405 二次性高血圧

That would be helpful, yes, and could be affected via
"data_source.lang".

> or, if searching on '401'
> 
>       401 Essential hypertension
>       401     Hypertension artérielle essentielle
>       401     原发性高血压
> 
> and when I am searching to attach a code in the same praxis I want to see
> not the duplicated list above but
> 
>       401 Essential hypertension
>       402 Hypertensive heart disease
>       403 Hypertensive renal disease
>       404 Hypertensive heart and renal disease
>       405 Secondary hypertension
> 
> and when I look at her patient's coded issues, must I see
> 
>       2506    Diabète avec troubles circulatoires périphériques
>       3651    Glaucome à angle ouvert
>       401     Hypertension artérielle essentielle
>       78039   Crise convulsive NS
>       7893    Gonflement abdominal ou pelvien, masse ou bosse

There's no reason why you should *have* to see the French version.

But that got nothing much to do with how the reference lists of
codes are stored in the backend - in particular, the way they are
stored now readily allows for what is desired here.

Back when I designed the coding I asked for input
and got little feedback.

> Likewise if this patient were to need care outside the praxis, to be able
> to export into the PDF the codes in a language of choice. I expect this
> would be easier if the language-translated terms had been normalized. 

You seem to assume that instances of coding systems are exchangeably
compatible across languages ?

This is the case less often than one might think.

I very much think that it'll be a lot more useful to export the
actual clinical label along with the code/coding system/version thereof
a clinician previously attached to said label.

Karsten
-- 
Empfehlen Sie GMX DSL Ihren Freunden und Bekannten und wir
belohnen Sie mit bis zu 50,- Euro! https://freundschaftswerbung.gmx.de



reply via email to

[Prev in Thread] Current Thread [Next in Thread]