[Gnumed-devel] Re: FreeDiams' LK_MOL_ATC

[Jim Busser]

On 2010-07-28, at 10:10 AM, Eric MAEKER wrote:

>> Also it may be that within any country
>> 
>> 1) the molecule name might not match what other countries use even in the 
>> same language (e.g. maybe they kept their own historical molecule names 
>> instead of updating data files to INN)
>> but
>> 2) within any one country data source, each molecule likely has a distinct 
>> name

do we have agreement on the above?

>> and so what about a table
>> 
>>      ATC_WITH_NAMES
>> 
>> with columns
>> 
>>      ATC
>>      INN_EN
>>      INN_FR
>>      INN_DE
>>      ...
>>      MOL_NM_CA (which may be the same or different from IN_EN)
>>      MOL_NM_US (which may be the same or different from IN_EN)

do we have agreement that the above has may some value in keeping together the 
data (at most, one record per ATC)?

Even if the sources (automated and manual) would be kept separate there is IMO 
value to the above, no?


>> and noting only that in the case of ATC for combination medications (whose 
>> INN_EN contains "and"), the country sources often provide no combined MOL_NM 
>> as the countries provide only the *component* molecule names. But despite 
>> that Canada provides an ATC for its combination drugs, we do not really need 
>> [what would be] the__ [associated multi-molecule] INNs except for 
>> convenience of information... we do not depend on the ATC [or INN] of the 
>> combination [name] to calculate interactions or allergy. Correct?
> 
> I do not understand here...

In the case of multi-molecule drugs, there does not (in FD) exist in the drug 
table any name beyond the commercial or brand name... the drugs contains no 
"multi-molecule" name. So, from the government data sources, we need only the 
commercial name, and in a separate table the molecule names (and we accept the 
ATCs if these are helpfully provided).

We do not depend on the source to provide a "multi-molecule name".

Despite that there exists in the WHO an ATC which (for multi-molecule drugs) 
*does* specify a "multi-molecule INN", we do not need it. We can include it in 
our reference tables, and even in the tooltips, but we do not in FD *depend* on 
such multi-molecule INNs for the interactions or allergy-checking. Is this 
correct?

-- Jim