[Top][All Lists]

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

Re: [Gnumed-devel] clin_medication

From: Hilmar Berger
Subject: Re: [Gnumed-devel] clin_medication
Date: Wed, 20 Oct 2004 21:03:33 +0200

On Wed, 20 Oct 2004 09:14:49 +0200
Hi all,
Karsten Hilbert <address@hidden> wrote:
> > Also I found a drug database hat has been open sourced. How many 
> > options do we now have to choose from?
> >
> Well, the problem of drug databases isn't their schema. Any
> schema can have a driver written for it. The problem lies with
> the content. And with having some mechanism to ascertain
> oneself of the reliability of the content.

Maybe we can use something like a local and permanent cache of all drugs one or 
more doctor is using. I.e. we don't store the details of the prescribed drug in 
clin_medication. Instead something like a "minimal defining drug data set" is 
stored in some other table, caching them for further use. This data will never 
be deleted as long as a particular item is referenced from within 
clin_medication. The minimal drug data set contains or links to any number of 
drug databases, where additional information can be retrieved as long the 
reference to the drug database is valid. When the link one day is no longer 
valid, you know the drug from the cache entry. When you choose a drug from some 
drug database on prescription, the cache is checked for presence of this 
particular item. If present, the item from cache is referenced. If not, a new 
cache entry is added and referenced. 
Given the fact that usually doctors will use only a limited number of drugs, 
cache size should stay within reasonable limits.
This would solve the problem of duplicated drug information in clin_medication 
and make sure you have always enough information to reidentify a particular 
drug, independent of existence of other sources.


reply via email to

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